The (: )

HORTON H. HOBBS, JR. and C.W. HART, JR.

SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY • NUMBER 364 SERIES PUBLICATIONS OF THE SMITHSONIAN INSTITUTION

Emphasis upon publication as a means of "diffusing knowledge" was expressed by the first Secretary of the Smithsonian. In his formal plan for the Institution, Joseph Henry outlined a program that included the following statement: "It is proposed to publish a series of reports, giving an account of the new discoveries in science, and of the changes made from year to year in all branches of knowledge." This theme of basic research has been adhered to through the years by thousands of titles issued in series publications under the Smithsonian imprint, commencing with Smithsonian Contributions to Knowledge in 1848 and continuing with the following active series: Smithsonian Contributions to Anthropology Smithsonian Contributions to Astrophysics Smithsonian Contributions to Botany Smithsonian Contributions to the Earth Sciences Smithsonian Contributions to the Marine Sciences Smithsonian Contributions to Paleobiology Smithsonian Contributions to Zoo/ogy Smithsonian Studies in Air and Space Smithsonian Studies in History and Technology In these series, the Institution publishes small papers and full-scale monographs that report the research and collections of its various museums and bureaux or of professional colleagues in the world of science and scholarship. The publications are distributed by mailing lists to libraries, universities, and similar institutions throughout the world. Papers or monographs submitted for series publication are received by the Smithsonian Institution Press, subject to its own review for format and style, only through departments of the various Smithsonian museums or bureaux, where the manuscripts are given substantive review. Press requirements for manuscript and art preparation are outlined on the inside back cover.

S. Dillon Ripley Secretary Smithsonian Institution SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY • NUMBER 364

The Shrimp Genus Atya (Decapoda: Atyidae)

Horton H. Hobbs, Jr. andC.W. Hart, Jr.

SMITHSONIAN INSTITUTION PRESS City of Washington 1982 ABSTRACT

Hobbs, Horton H., Jr., and C.W. Hart, Jr. The Shrimp Genus Atya (Deca- poda: Atyidae). Smithsonian Contributions to Zoology, number 364, 143 pages, 53 figures, 1 table, 1982.—The genus Atya, as characterized in this study, com- prises an assemblage of 11 tropical freshwater ranging through the Antilles and along the Atlantic and Pacific versants of Middle and South America and in western Africa. The literature related to these is reviewed, and the taxonomic characters are discussed. Correlations of carapace length and total length are presented, and ecological and life history data reviewed. A discussion of relationships and dispersal is followed by a definition of the genus, its range, and a key to the species. For each of the latter, a complete synonomy, a review of the literature, a summary of available illustrations, a diagnosis, an illustrated description, the range, a list of localities and the specimens examined, observed variations, its ecological distribution, and life history notes are provided. Atya brachyrhinus, known to occur in only a single cave on Barbados, West Indies, is the only new taxon introduced. Four species are largely restricted to the Pacific slope of the Americas: Atya crassa, A. dresslen, A. margantacea, and A. ortmannwides; two are confined to the Antilles: A. brachyrhinus and A. lanipes; two to Africa: A. afncana and A. intermedia. Two species occur in the Atlantic versant of both continents: A. gabonensis and A. scabra; and only one, A. innocous, is widespread on both slopes of Middle America, although A. crassa and A. scabra are known to occur in a few localities on the Atlantic and Pacific slopes, respectively.

OFFICIAL PUBLICATION DATE is handstamped in a limited number of initial copies and is recorded in the Institution's annual report, Smithsonian Year. SERIES COVER DESIGN: The coral Montastrea cavernosa (Linnaeus).

Library of Congress Cataloging in Publication Data Hobbs, Horton Holcombe, Jr., 1914- The Shrimp genus Atya (Decapoda: Atyidae) (Smithsonian contributions to zoology ; no. 364) Bibliography: p. 1. Atya—Classification. 2. Crustacea—Classification. I. Hart, C.W Ir II Title III Series. QL1.S54 no. 364 [QL444.M33] 591s 82-800 |595.3'843] AACR2 Contents

Page Introduction 1 Acknowledgments 1 Collections Studied 2 A Brief History of the Genus Atya 2 Composition of the Genus Atya as Defined Herein 4 Taxonomic Characters 4 Carapace 4 Abdomen 5 Sternum of Fifth Abdominal Segment 5 Sternum of Sixth Abdominal Segment 5 Telson 6 Preanal Carina 7 Ocellar Beak 7 Antennule 7 Antenna 7 Gnathal Appendages 7 First and Second Pereiopods 9 Third Pereiopod 10 Fourth and Fifth Pereiopods 11 First and Second Pleopods 11 Branchial Complement 11 Remarks 13 Measurements and Size Correlations 14 Habitat 14 Life History 15 Relationships and Dispersal 17 Characters Compared 20 Genus Atya (sensu stricto) 22 Key to Species of the Genus Atya 26 Atya afneana Bouvier 27 Atya brachyrhinus, new species 35 Atya crassa (Smith) 39 Atya dresslen Abele 44 Giebel 48 Atya innocous (Herbst) 56 Atya intermedia Bouvier 76 Holthuis 82 Atya margantacea A. Milne-Edwards 91 Atya ortmannwides Villalobos 101 (Leach) 106 Literature Cited 133

The Shrimp Genus Atya (Decapoda: Atyidae)

Horton H. Hobbs, Jr. and C.W. Hart Jr.

Introduction localities, and enumeration of specimens exam- ined. A discussion of variations, what is known of The present study was prompted by the receipt the ecological distribution and life history, occa- of a collection of shrimps from containing sional remarks, and illustrations conclude the specimens of Atya gabonensis Glcbe\, 1875, that lent sections devoted to each. The common names credence to the report of Koelbel, 1884, that his and economic importance of certain members of Evatya sculptilts (a subjective junior synonym of A the genus are pointed our under the treatments gabonensis) indeed had been collected in the Ori- of Atya innocous (Herbst, 1764) and Atya scabra noco River. In preparing an announcement of the (Leach, 1815). rediscovery of this shrimp on the South American ACKNOWLEDGMENTS.—Many persons have as- continent, one of us (Hobbs) reviewed the most sisted us in this study that has elicited the greater readily available information concerning the part of our attention over a period of some two American members of the genus Atya and in so years. We are most grateful to those individuals doing became aware of the need for both bringing who have lent specimens to us, many collections together what is known about these shrimps and representing previously unpublished locality rec- for a critical study of the species that have been ords: Lawrence G. Abele of Florida State Univer- assigned to the genus. sity; Jorge A. Cabrera J. of the Instituto de This summary of the present state of our knowl- Biologia, Universidad Nacional Autonoma de edge of the larger tropical African and American Mexico; Jacques Forest of the Museum National atyids includes as exhaustive a bibliography as d'Histoire Naturelle, Paris; Willard D. Hartman we have been able to assemble. Previous work is of the Peabody Museum, Yale University; Lipke reviewed, an analysis of taxonomic characters B. Holthuis of the Rijksmuseum van Natuurlijke employed in recognizing the several species is Historie, Leiden; Raymond W. Ingle of the Brit- offered, and a discussion of relationships and ish Museum (Natural History); Catharine Kessler distribution prefaces a synonomy and definition and Herbert VV. Levi of the Museum of Compar- of the genus. This is followed by a key to the 11 ative Zoology, Harvard University; C. Bruce recognized species. Diagnoses and descriptions of Powell of the University of Port Harcourt. Ni- each precede a statement of the range, list of geria; and Alejandro Villalobos F. of the Univ- ersidad Autonoma Metropolitana Iztapalapa, Horton H. Hobbs, Jr., and C. W. Hart, Jr., Department of Invertebrate Mexico. The courtesies extended to one of us Zoology, National Museum of Natural History, Smithsonian Institu- tion, Washington, D.C. 20560. (Hart) by Drs. Forest, Holthuis, and Ingle during SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY visits to their respective museums are also appre- Library in obtaining a number of references that ciated. For lending us the atyid section of his are included in the "Literature Cited" herein. For manuscript treating the shrimps of the Philip- her help in ferreting errors in references through- pines and Indonesia, we are grateful to Fenner A. out the text, we are grateful to Georgia B. Hobbs. Chace, Jr., of the Smithsonian Institution, as we Finally, we extend our thanks to S. Dillon are for his criticisms of the penultimate draft of Ripley, Secretary of the Smithsonian Institution, this paper. For information concerning types, we who provided financial aid that made possible acknowledge with gratitude the assistance of visits to the museums noted above. H.-E. Gruner of the Zoologisches Museum, Ber- COLLECTIONS STUDIED.—Most of the materials lin; Charlotte Holmquist of the Naturhistoriska mentioned or that were examined during this Rijksmuseet, Stockholm; and Gerhard Pretz- study are in the collections of the following mu- mann of the Naturhistorisches Museum, Vienna. seums or persons. We also extend our thanks to Bruce E. Felgen- BM British Museum (Natural History), London, Eng- hauer of Florida State University for lending us land color photographs of Atya margaritacea and A. in- IBM Instituto de Biologia de la Universidad Nacional nocous from Panama, for observations he had Autonoma de Mexico, Mexico, D.F., Mexico LGA collection of Lawrence G. Abele, Florida State made on the color of these two shrimps, and for University, Tallahassee, Florida, USA (recently forwarding to us a copy of his unpublished man- donated to USNM) uscript, coauthored with Dr. Abele, reporting MCZ Museum of Comparative Zoology, Harvard Uni- observations on the mating behavior of the latter versity, Cambridge, Massachusetts, USA shrimp. MHNP Museum National d'Histoire Naturelle, Paris, For their advice and assistance in statistical France NMW Naturhistorisches Museum Wien, Vienna, Austria analyses, we thank J.W. Craig and Lee-Ann PM Peabody Museum of Natural History, Yale Uni- Hayek of the Smithsonian and J.L. Russo of the versity, New Haven, Connecticut, USA National Marine Fisheries Service. The help of RNHL Rijksmuseum van Natuurlijke Historie, Leiden, Janice Clark in obtaining measurements and as- The Netherlands sisting in the preparation of the maps and of SMNH Naturhistoriska Riksmuseet, Stockholm, Sweden USNM former United States National Museum collec- Carolyn B. Gast, who inked the illustrations in tions deposited in the National Museum of Figures 2, 3, and 5, is gratefully acknowledged. Natural History, Smithsonian Institution, We are deeply indebted to Lipke B. Holthuis Washington, D.C. not only for criticisms of the manuscript but also ZBM Zoologisches Museum der Humboldt-Universitat, for his having contributed many of the biblio- Berlin, East Germany graphic references of which we were unaware and for providing us with the citation of Gerstaecker and Ortmann (1891-1901), collating the various sections therein with dates of publication. We A Brief History of the Genus Atya acknowledge his invaluable assistance with gen- Although the literature dealing with the indi- uine appreciation. For their suggestions and crit- vidual species is discussed for each, a brief account icisms of the manuscript, we also thank Lawrence of the attention that has been accorded the mem- G. Abele, C. Bruce Powell (who assisted us with bers of the genus seems appropriate here. The much of the literature pertaining to the African earliest report of the existence of a species be- Atya), and the following Smithsonian colleagues: longing to the genus Atya that we have encoun- Margaret A. Daniel, Raymond B. Manning, and tered is that of Marcgrave (1648), who presented Isabel Perez Farfante. Ms. Daniel deserves special an illustration and a brief description of thanks for her varied assistance throughout the "Guaricuru" in Brazil, a common name that was course of this study. We acknowledge with appre- identified with Atya scabra by Lemos de Castro ciation the indispensable assistance of Carolyn S. (1962). Marcgrave's description (p. 187) was as Hahn and Jack F. Marquardt of the Smithsonian follows. NUMBER 364

Guaricuru Brasiliensibus, Gammarus vulgo Camaraon; qua- (1898). Thus by 1900, 13 species and one varietal tuor digitos longus: crura habet sex, tribus internodiis con- name had been applied to the shrimps herein stantia, cum unguiculo in extremitate: anterius par crurum recognized as belonging to the genus Atya. quasi suras habet, crassius relinquis & vicem brachiorum fungens; paulo plus duobus digitis longum: medium parses Bouvier's (1904) study of shrimps belonging to quidigitum longum, ultimum paulo plus uno, omnia acu- the genus included the descriptions of two new leata seu spinosa. Cirrhos prope os prodeuntes habet sex; members from Africa: Atya africana from the duos longissimos, nimirum tres digitos singulos, versus pos- Ogooue River, Gabon, and A. intermedia from Sao teriora reflexos; duos unum digitum, & duos unum digitum, Tome. Most of the above were reviewed in Bou- & duos fernidigiturn: sub ore parva tenacula quatuor cras- siuscula, & quatuor tenuiora, quibus cibum tenet. Oculi vier's monograph (1925); the synonyms, exclud- fimiles reliquis. Color totius fuscus. Cocti eduntur. ing misidentifications, recognized by him are in- cluded in the following list. Not until Atya ortman- nioides was described from the Pacific versant of Mexico by Villalobos (1956) was additional in- The notice of a second shrimp that is currently formation added to our knowledge of the com- assigned to the genus Atya was that of Gronovius position of the genus, and about seven years later (1764), who described and illustrated "Astacus the presence of Atya lanipes on Saint Thomas, 988" from "Oceano Americano ad Martinicam." Virgin Islands, was made known by Holthuis This shrimp was later designated "Astacus Na- (1963). The latter (1966) presented a noteworthy soscopus" by Meuschen (1778) and "Cancer (As- analysis of the of several of the species, tacus) Innocous" by Herbst (1792)—descriptions comparing A. scabra with A. sulcatipes and A. based on the specimen described by Gronovius rivalis. He declared "Atya margaritaria clavipes" a (Meuschen's publication was declared invalid in synonym of A. sulcatipes, restricted the name A. Opinion 260 of the International Commission on scabra to "the East American species," and rec- Zoological Nomenclature (1954:267)). ognized A. mexicana, A. margaritacea, and A. punctata Leach (1815) introduced the name Atys scaber as synonyms of Atya scabra. Atya intermedia was for a shrimp from an unknown locality, emending contrasted with A. sulcatipes, and he noted the it to Atya scabra the following year. The name Atya close affinity of the former with A. innocous Herbst. mexicana was proposed for a shrimp from Mexico The validation of this name and Holthuis' clari- by Wiegmann (1836), and Newport (1847) de- fication of its synonomy constituted a particularly scribed Atya sulcatipes from the Cape Verde Is- valuable contribution. Although in 1966 he did lands and Atya occidentalis from . Some 17 not mention A. lanipes, A. crassa, and A. gabonensis years later, A. Milne-Edwards (1864) applied the or present formal synonymical lists, his concept names Atya margaritacea and A. robusta to specimens of the existing species of larger atyids seems to that were thought to have been collected on New have been as noted in the following list. The only Caledonia. Shrimps that had been obtained from species described since the publication of Hol- streams on the Pacific slope in Guatemala pro- thuis' just-mentioned contribution is Atya dressleri vided the material on which Smith (1871) named from Panama by Abele (1975). The species and Evatya crassa, Atya rivalis, and Atya tenella. A shrimp synonyms recognized herein are noted in the list from Gabon was designated Atya gabonensis by below. Giebel (1875), and three years later Kingsley Other important contributions to our knowl- (1878a) described Atya punctata from Haiti. An edge of members of the genus are the morpholog- ornate member of the genus from "the Orinoco" ical study of Atya scabra by Villalobos (1943), the was made known by Koelbel (1884), who desig- observations and illustrations of Chace and nated it Euatya sculptilis. Ortmann (1890) applied Hobbs (1969), the studies on the larval stages of the name Atya sculptata to a shrimp from western A. lanipes and A. innocous by Hunte (1975, 1977, Africa, and two specimens from Cameroon were 1979b), and the work on functional morphology designated a variety of A. Milne-Edwards' spe- and ecology of A. innocous and A. scabra by Fryer cies, Atya margaritacea var. claviger, by Aurivillius (1977). SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

COMPOSITION OF THE GENUS Atya As DEFINED HEREIN

(Shrimp names enclosed in parentheses signify synonyms of the species under which they are listed)

BOUVIER (1925) HOLTHUIS (1966) PROPOSED HEREIN Atya africana Bouvier, 1904 Atya africana Atya africana Atya intermedia Bouvier, 1904 Atya intermedia Atya intermedia Atya innocous Atya innocous (Herbst, 1792) Atya robusta A. Milne-Edwards, 1864 {Atya robusta) {Atya robusta) Atya occidentalis Newport, 1847 {Atya occidentalis) {Atya occidentalis) {Atya lenella Smith, 1871) Atya tenella {Atya tenella) Atya scabra (Leach, 1815) Atya scabra Atya scabra {Atya mexicana YViegmann, 1836) {Atya mexicana) {Atya mexicana) {Atya punctata Kingsley, 1878a) {Atya punctata) {Atya punctata) {Atya sulcatipes) {Atya margaritacea claviger) {Atya margaritacea A. Milne-Edwards, {Atya margaritacea) Atya margaritacea 1864) {Atya sulcatipes Newport, 1847) Atya sulcatipes {Atya margaritacea var. claviger {Atya margarilaria Aurivillius, 1898) clavipes) {Atya rivalis Smith, 871) Atya rivalis {Atya rivalis) Atya gabonensis Giebel, 1875 not mentioned Atya gabonensis {Euatya sculptilis Koelbel, 1884) not mentioned {Euatya sculptilis) {Atya sculptata Ortmann, 1890) not mentioned {Atya sculptata) Evatya crassa Smith, 1871 not mentioned Atya crassa not mentioned Atya ortmannioides Villalobos, 1956 not mentioned Atya lanipes Holthuis, 1963 Atya dressleri Abele, 1975 Atya brachyrhinus, new species

Taxonomic Characters useful taxonomic features are the sculpture and gross texture. Only two of the recognized species Those features that have been found to be of exhibit conspicuously sculptured features: in Atya use in distinguishing between, or may be con- crassa there are ridges studded with spines over sistent among, the several species of the genus much of the cephalic region, and the correspond- Atya are briefly discussed in the following para- ing areas in A. gabonensis exhibit strong ridges that graphs. Except in secondary sexual characters lack spines. Larger individuals of A. scabra and A. there are only slight differences in the males and margaritacea bear rugose, somewhat weaker sculp- females of all members of the genus, and we have tured cephalolateral regions, but there is little noted few apparent differences in these features irregularity in the carapace of smaller individuals. between species. In A. scabra and A. margaritacea the carapace, CARAPACE.—Among the surface structures of especially the lateral regions, is studded with a the carapace that have been found to serve as comparatively dense pile of short stiff brown setae NUMBER 364 that may become abraded in later intermolt sleri the spine, if present, is usually much reduced stages, and is lacking or represented by very small in size although in the holotype it is moderately fine setae in the glabrous A. innocous and A. lanipes. strong. In collecting living specimens of A. scabra and A. ABDOMEN.—The presence or absence of corni- innocous on Dominica, it was discovered that one fied denticles on the ventral margins of the second need not look at the shrimp to recognize its through the fifth abdominal pleura aids in the identity, for if it could be held in one's hand for recognition of certain species. Their occurrence more than a few seconds it was indeed A. scabra. on the second is characteristic of most members The glabrous carapace of A. innocous makes mem- of Atya scabra, but occasional specimens from the bers of this species almost impossible to grasp for western Atlantic versant and all of those in the any length of time. eastern Atlantic watershed lack such spinules on The size and sometimes at least the number of that pleuron. Similar denticles are sometimes punctations on the carapace are highly variable. present on only the fifth pleuron in A. africana, Such is particularly noticeable in available spec- but some specimens exhibit none. In A. innocous, imens of A. scabra. Even in specimens from a the denticles are more delicate, and there is tre- single locality, the size and density of punctations mendous variation in the presence (number and differ markedly. Never, however, are they so con- conspicuousness) or absence on one or all of the spicuous in A. innocous or in A. lanipes and their third through fifth pleura. (See "Variations" un- allies as they frequently are in A. scabra. der Atya innocous.) Such denticles have not been The principal features of the rostrum that have observed in A. crassa, A. dressleri, A. gabonensis, A. proven to be useful in distinguishing between intermedia, A. lanipes, and A. ortmannioides. certain species are the presence or absence of a STERNUM OF FIFTH ABDOMINAL SEGMENT.—In dorsomedian row of spines, the elevation of the all members of the genus except Atya crassa and median carina, and the contour of the lateral A. gabonensis, the narrow sternal bar between the margins. The latter may taper rather gently from fifth pleopods bears a comparatively small, often the base to the apex, may be contracted abruptly laterally compressed, ventrally directed tubercle. in an angular or subangular bend, or may even In A. crassa, however, the tubercle is moderately be notched. As pointed out by Holthuis to strongly developed, and in A. gabonensis it is (1966:235), however, in juveniles this character is comparatively enormous, resembling a curved not always reliable, for angles that are well horn, the apical end of which presses against the marked or even subacute in the adult may be sternum of the sixth abdominal segment when broadly rounded or almost imperceptible in the "tail fan" is flexed. This prominence actually young individuals. The size and ornamentation prohibits complete flexion of the abdomen in this of the dorsal and ventral keels are subject to shrimp. considerable intraspecific variation as are the STERNUM OF SIXTH ABDOMINAL SEGMENT.—The span of the anterior extension, the subapical sternal element of this segment consists of a plate spines, and the depth of the sulci between the that covers almost the entire ventral surface of median and lateral carinae. The long acumen the segment; thus it spans the distance between (apical part of the rostrum) of Atya ortmannioides the pleura and extends from the posterior mem- furnishes perhaps the most ready character for branous part of the fifth sternal area to the ventral recognizing members of this species. base of the telson, bearing a posteromedian notch In all representatives of the genus except Atya that receives the base of the preanal carina. The dressleri and some members of A. ortmannioides and relative dimensions of this plate are useful in A. gabonensis, the pterygostomian angle is rather distinguishing between certain species (see Figure strongly produced, often in a distinct spine. In 1). The longer plates occur in A. lanipes and A. the latter it is not so strongly developed as in ortmannioides, and the shorter ones are typical of most individuals of other species, and in A. dres- A. crassa and A. gabonensis; in the remaining spe- SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

h

FIGURE 1.—Ventral view of preanal carina, sternum of sixth abdominal segment, and median tubercle (or projection) on fifth abdominal segment in: a, Atya lanipes; b, A. ortmannioides; c, A. brachyrhinus; d, A. innocous; e, A. tenella (= A. innocous); f, A. intermedia; g, A. dressleri; h, A. africana; i, A. margantacea; j, A. rivalis (= A. margaritacea); k, A. scabra; /, A. sulcatipes (= A. scabra); m, A. crassa; n, A. gabonensis. cies the length of the plate is intermediate be- ortmannioides 2.0 to 2.7, and A. scabra 1.6 to 1.9. tween these extremes. The numbers of spines on each side of the dorsal TELSON.—The ratio of the length to the width surface range from 7 to 9 in A. dressleri to no more of the telson is useful in separating certain species than 5 in A. africana, A. crassa, and A. gabonensis. from others. Among the specimens measured be- In the remaining species there are 5 or 6 in A. longing to several species, the ratios range from lanipes and A. margaritacea, 5 to 7 in A. scabra, and 1.2 to 3.0 and are dispersed as follows: Atya africana 5 to 8 in A. innocous, A. ortmannioides, and probably 1.8 to 1.9, A. crassa 1.2 to 1.4, A. dressleri 1.9 to also in A. brachyrhinus, of which only two speci- 3.0, A. gabonensis 1.2 to 1.5, A. innocous 1.9 to 2.2, mens are known. A. lanipes 2.1 to 2.6, A. margaritacea 1.6 to 1.8, A. Except in injured specimens the caudal margin NUMBER 364 of the telson is uniformly provided with two pairs species, and the occurrence of such spinules, an- of posterolateral spines situated mesial to the gular rostral margins, and a row of contiguous posterolateral angles; the more mesial pair is scales on the flexor surface of the propodus of the longer. Between these spines is a row of strong third pereiopod are peculiar to A. scabra. Spinules plumose setae flanked dorsally by a row of much unaccompanied by tapering rostral margins or a finer ones, and there is a dorsomedian tubercle linear series of scales on the flexor surface of the that may or may not reach or overreach the comparatively short propodus of the third pereio- caudal margin of the telson. pod are characteristic of A. gabonensis. The subli- PREANAL CARINA.—Details of the form of the near arrangement of the spinules on the lateral preanal carina are highly variable, but despite part of the dorsal surface of the penultimate the variability its form is diagnostic for at least podomere of the antennule in A. margaritacea usu- three species. The more distinctive ones are those ally serves as an accessory characteristic that may of Atya lanipes (in which the spine is sometimes be used in distinguishing this shrimp from the absent or vestigial), A. gabonensis (in which it is closely allied Atya scabra. Useful also are the num- subconical and directed almost ventrally), and A. bers of spinules on the dorsal surface of the distal intermedia (in which there exist two posteriorly two podomeres of the peduncle as are those in the directed spines). In the other species the carina is distal rows of all three podomeres. less distinctive, although it is subconical in A. To some degree, the ratio of length to width of crassa. the three podomeres of the antennule, especially OCELLAR BEAK.—This is a median prominence the penultimate, provides an additional feature situated between the eyes that in the alpheids was that aids in the recognition of certain species, but designated the "bee ocellaire" by Coutiere the range of intra- and interpopulational varia- (1899:108); it was referred to as the "carene tion is rather broad. The attenuation of all of the antennulaire" by Bouvier (1925:23). In all of the pereiopods in A. lanipes is conspicuous, and this is members of the genus Atya except A. gabonensis, it decidedly apparent in both the peduncle and is comparatively inconspicuous (concealed by the flagella of the antennule. As compared with A. rostrum above, the eyes to the sides, and the scabra, for example, the proximal thickened part antennules below). In A. gabonensis, however, it is of the lateral flagellum is longer and is less mark- produced in a dorsally arched blade that ends in edly set off from the distal more slender section. an acute apex reaching almost as far anteriorly ANTENNA.—We have found no feature of the as the tip of the stylocerite. antenna except possibly the lateral spine on the ANTENNULE.—The size and the intensity of cor- scaphocerite that might be useful in distinguish- nification of the spinules, or denticles, on the ing between the various species of the genus. We dorsal surface of the three articles of the anten- have not attempted to quantify the relative de- nular peduncle are in part reflections of both the gree of development in the several species, but it size of the individual and the stage in the molting seems to be much more prominent in some than cycle, the dark coloration intensifying during in- in others. termolt stages. Except in small individuals their GNATHAL APPENDAGES.—To provide an appre- distribution or absence (especially on the dorsal ciation of the structure of the gnathal append- surface of the basal segment proximal to the distal ages, we have chosen to illustrate those of Atya marginal row) in combination with other features innocous (Figure 2; see also Fryer, 1977), which aids in recognizing certain species. For example, seem as generalized as any. Minor differences, premarginal spinules occur typically on the dorsal chiefly in the development of setae, seem to exist surface of the proximal segment in Atya africana, between some of the species, although none that A. gabonensis, A. scabra, and sometimes in A. mar- we have noted are any more striking than that garitacea. The combination of this feature with exhibited by the shape of the flagellar lobe of the tapering rostral margins is unique to the former first maxilliped in A. innocous and A. gabonensis (cf. SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 2.—Appendages of Atya innocous (all from male, Mannet's Gutter, Dominica): a, lateral and submesial views of mandible; b, first maxilla; c, second maxilla; d, first maxilliped; e, second maxilliped;/, third maxilliped, g, first pereiopod; h, second pereiopod. (Scales marked in 1 mm increments.) NUMBER 364

y

FIGURE 3.—Gnathal appendages of male of Atya gabonensis from Volta River, : a, first maxilla; b, first maxilliped. (Scales marked in 1 mm increments.)

Figures 2d and 36). The relative length of the us, and we are puzzled by the absence of setae exopod of the third maxilliped in comparison to that bear scraping denticles in some specimens of that of the endopod seems to be consistent, as those species that often, if not usually, exhibit does the range of numbers of bands of setae on them. the distal two podomeres; however, we have not In his study of functional morphology in the examined a sufficiently large series to be assured atyids of Dominica, Fryer (1977:58) pointed out that our preliminary observations are valid. that some of the terminal "bristles [of the cheli- FIRST AND SECOND PEREIOPODS.—There is re- peds] of A. innocous are armed distally with minute markable uniformity in the structure of the first denticles . .. that facilitate scraping and sweep- and second pereiopods in all members of the ing: no such are present in A. scabra. The differ- genus. Minor variations have been noted in the ence is related to the relative importance of scrap- form, length, and perhaps density of the setal ing in the two species: A. innocous scrapes fre- tufts capping the distal extremities of the fingers, quently, A. scabra seldom." The setae of the chelae but to what extent these differences are governed of the remaining species recognized herein have by the stage of the molting cycle is not clear to been examined, and the denticulate setal type has 10 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY been found only in A. innocous and A. lanipes. loped, and the anterolateral angle of this podo- Villalobos (1956:466) also found them in A. ort- mere on both the third and fourth pereiopods mannioides. On the basis of our limited examina- each bears a heavy spine. Another useful feature tions, such setae cannot be said to be typical of of the coxa is the degree of development or ab- all individuals of any of the three species, for we sence of a posteromedian prominence bearing failed to find them in a number of specimens of setal tufts. Whereas it is strongly developed in A. A. innocous from several localities on Dominica, crassa and A. gabonensis, moderately so in A. mar- and they were not evident in our preparations of garitacea and A. scabra, it is absent in A. dressleri, A. samples from A. ortmannioides. When they are lanipes, and A. ortmannioides. Equally as useful are present, however, there is reason to believe that features of one or more of the distal four podo- the shape of the denticles may well serve as meres. For example, the ratios of their length to diagnostic features, as they seem to be for A. width serve readily to distinguish Atya lanipes innocous and A. lanipes (cf. Figure 4a,b), (merus) and A. crassa (propodus) from other spe- THIRD PEREIOPOD.—Among the most useful cies. The absence of tubercles (sometimes few characters for distinguishing between the several very weak ones present) or spines on the lateral species are those associated with the third pereio- surface of the merus is characteristic of A. lanipes, pod. The coxa of Atya crassa is unique in two and the presence or absence of strongly sclerotized respects: the ventrodistal margin is distinctly scal- extremities of spines or scalelike tubercles aids in

FIGURE 4.—Bristles from distal tufts of first pereiopod: a, distal part of bristle bearing scraping denticles in Atya innocous; b, same in A. lanipes; c, distal part of bristle that serves as a brush seta in A. lanipes. (Scales marked in 1 mm increments.) NUMBER 364 11 separating certain species from others. The row uncertain as to whether or not the merus of the of contiguous or subcontiguous, strongly cornified third and fourth pereiopods of small individuals "scales" or spines on the mesioflexor surface of of all species bears a row of three ventral spines the propodus distinguishes A. africana and A. scabra and one distal ventrolateral spine. Evidence ex- from the remaining members of the genus. The ists, however, that all have fewer with increasing arrangement of denticles and/or spines on the size of body. Apparently there is not a direct flexor surface of the dactyl is also helpful: as intraspecific correlation between the number of pointed out above, the denticles (sometimes only persisting spines and the size of the shrimp. one, or none) are limited to a small cluster at the The spinules on the flexorsurfac e of the dactyl base of the corneous tip of the dactyl in A. crassa of the fifth pereiopod of A. ortmannioides (Figure and A. gabonensis; in three species they are ar- 43c) form a pectinate row that has been observed ranged in a single row (A. africana, A. margaritacea, infrequently in other species. A comparative and A. scabra); in the remaining ones they are study of this podomere in a series of specimens of dispersed in two, sometimes irregular, rows. each species might prove rewarding. The quantity and degree of development of FIRST AND SECOND PLEOPODS.—In these ap- pubescence on the pereiopods are variable as pendages there is less variation between those of pointed out by Chace and Hobbs (1969:63), who different species than between those of females of noted that in none of their specimens of Atya the same species at different stages of their adult lanipes from Puetro Rico are the pereiopods life. There is evidence that in Atya, as in Macro- "clothed in hair dense enough to conceal the brachium rosenbergii (De Man) as pointed out by underlying surface" as they are in those from the Nagamine and Knight (1980:147, 148, fig. 1), the type-locality. Hunte (1975:68) also observed that prepartureal molt enhances the setation of the among his Jamaican specimens "the intensity and pleopods (cf. Figure bd and 5e,f). This is accom- distribution of pubescence on the last three pe- plished largely by the addition of setae designated reiopods is subject to much variation ...." The as "ovigerous" or "temporary" setae by Naga- presence of a conspicuous oblique row of tufts of mine and Knight. In Atya, the molt following the long plumose setae on the lateral surface of the ovigerous condition seems to result in a reduction merus has been observed in specimens of several of the ovigerous setae, returning the pleopods to species and are distinctly most obvious in individ- a condition comparable to that existing prior to uals that have recently molted. Nevertheless the the prepartureal molt. Of course, many of the shaggy setal clusters on the third pereiopod of ovigerous setae disappear with the hatching of recently molted specimens of A. crassa seem to be the eggs. unique. We found no conspicuous differences in these The latter species is also unique in lacking a appendages in the males. There is considerable movable articulation between the propodus and variation in the spination of the appendix mas- dactyl of this appendage. culina, but except that it is more elongate in A. FOURTH AND FIFTH PEREIOPODS.—Further crassa than in the other species no unique features study should be accorded the fourth and fifth in any of the species were apparent to us. pereiopods. The series of movable spines on the BRANCHIAL COMPLEMENT.—The number and ventral and ventrolateral surfaces of the merus arrangement of the gills as reported for members and carpus seems to be more numerous, if not of the genus Atya (sensu stricto) are, for the most proportionately better developed, in the young part, consistent (Figure 6). Apparent variations and tends to be lost (the more proximal ones on have been noted, however, in elements associated all podomeres disappearing first) as the with the second and third maxillipeds. Holthuis increases in size. Because of inadequate samples (1966:233) reported the presence of a pleuro- in which series of stages are represented, we are branch on the body segment bearing the third 12 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

•

FIGURE 5.—Appendages of Atya innocous (all from Mannet's Gutter, Dominica); a, first pleopod of male; b, second pleopod of male; c, first pleopod of female; d, first pleopod of ovigerous female with ovigerous setae; e, basal part of third pleopod of nonovigerous female; /, third pleopod of ovigerous female. (Scales marked in 1 mm increments.) NUMBER 364 13

FIGURE 6.—Branchial apparatus ofleft side of Atya irmocous (epipods black). maxilliped in Atya scabra and A. sulcatipes, and A. rivalis in that a close-set row of "squamae" Abcle (1975:56) noted that one of the differences occurs on the ventral side of the third pereiopod. between A. dressleri and A. lanipes is in the absence Chace and Hobbs (1969) pointed out and illus- in the latter of a podobranch on segment VIII. trated the "denticles" on the ventral surface of Whereas these observations may well reflect vari- the second pleuron in A. scabra that are absent in ations in the three species, all of the members of A. rivalis (= A. margaritacea) and thus aid in the genus that have been examined in the present separating the two. To our knowledge, however, study exhibit the following branchial apparatus nothing concerning the close set row of tubercles and exopods on body segments VII through XIV on the third pereiopod has been recorded. Having shown in the accompanying tabulation. recalled Smalley's mentioning them, we found Branchial Apparatus and Exopods that this row, located on the flexor surface of the propodus, indeed serves not only to distinguish A. VII VIII IX X XI XII XIII XIV pleurobranch 1 1 1 1 1 scabra from Smith's A. rivalis but also from any arthrobranch 2 1 member of the genus except A. qfricana, and the podobranch 1 difference in the "form" of the tubercles in the mastigobranch 1 1 1 1 1 latter two is distinctive. Furthermore, this char- epipod 1 1 1 1 1 acter becomes evident in juveniles of both sexes exopod 1 1 1 and persists in both (except perhaps in regener- REMARKS.—In the early stages of the present ated appendages) until death of the . study one of the problems that we encountered Fortunately, the syntypes of A. margaritacea are was determining whether or not Atya rivalisdiffer s extant, and one of us (Hart) examined them in from A. scabra. Some years ago, Alfred E. Smalley, the Museum National d'Histoire Naturelle of Tulane University, while on a visit to the (Paris). When it was discovered that the linear Smithsonian Institution, told one of us (Hobbs) arrangement of the tubercles on the propodus that large males of Atya scabra differ from those of was lacking, one of the specimens was borrowed 14 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY and subjected to further study by both of us. Measurements of the carapace length and total Comparisons of the borrowed syntypic male with length were recorded for a series of 102 specimens specimens of Smith's Atya rivalis revealed no sig- of Atya innocous; these were plotted and subjected nificant differences. Thus in our opinion, the to analyses that provided us with the following types 6f \A. margaritacea, which were thought to equations: have been collected on New Caledonia, where no total length member of the genus (as defined here) occurs, = 3.41925 + 3.03262 (carapace length) were actually obtained from one or more streams carapace length on the Pacific slope of Central or South America. - -0.77307 + 0.32241 (total length) As we have noted elsewhere herein, the name Atya rivalis is a junior synonym of Atya margaritacea A. The regression lines y - 3.41925 + 3.03262* Milne-Edwards. On the basis of collections avail- and y = -0.77307 + 0.3224\x are shown in Figure la and b, respectively. The correlation (r) able to us, this shrimp appears to be limited to 2 the Pacific slope of Middle and South America. value of these statistics was 0.98881 (R » Except in a few localities, A. scabra does not occur 0.97775); thus 98% of the variability in either on the Pacific slope of the Americas but is wide- measurement is reflected in that of the other. spread on the islands of the Gulf of Mexico- Plotting the data available from all of the other , on the eastern slope of Mexico, and species of the genus showed that the variations in on the Atlantic versant of both South America them fall within the limits exhibited by A. inno- and Africa. cous. Specimens of a second potential species pair, Atya innocous and A. tenella, reported to occur on the Atlantic and Pacific slopes of Middle America Habitat have been compared, arid although some repre- On the basis of field experience in the West sentatives of the two can' be distinguished, the Indies and Mexico, published ecological notes, ranges of variation in the few characters in which the few habitat data accompanying collections, they differ are so great and overlap to such an and the apparent absence of populations in cer- extent even within a local population that in our tain areas within the range of the genus, we opinion the two must be considered to be conspe- suggest the possibility that perhaps oxygen con- cific. centration is the most important physical aspect Measurements and Size Correlations of the environment that determines the presence or absence of Atya. Most collections of these Because several previous authors have referred shrimps have been obtained in rapidly flowing to the sizes of their specimens by recording their streams, and the presence of rocks on the substrate total length (presumably the distance from the seems to be associated with the larger populations tip of the rostrum to the posterior extremity of (Figure 8). In the absence of rocks, debris and the telson), and all of our references to size are macrophytes provide adequate cover to support given in terms of the carapace length (the distance the presence of representatives of at least some from 'the posterior margin ofrthe orbit to the species. Streams within the range of the genus in midposterior margin of the carapace), we endea- which riffleso r segments of rapid flow are absent vored to determine whether there exists a close or situated some distance inland (that is, flowing enough correlation between the two measure- across an extended course of low gradient) seem ments to estimate one from the other. Further, we to lack populations of Atya. The absence of records also wished to know whether or not the relation- in streams traversing the low-lying coastal areas ship between the two is the same for one or more within the range of the genus tends to support the species. conclusion that these shrimps are largely re- NUMBER 364 15

10 15 20 25 30 35 40 10 20 30 40 50 60 70 80 90 100 110

Total Length Carapace Length

FIGURE 7.—Correlations of carapace length and total length in Atya innocous. (See "Measure- ments and Size Correlations" for explanation.) stricted in their distribution to rapidly flowing or and the studies of Fryer (1977) and Hunte swift segments of streams. Such areas must lie not (1979b). The adults frequent streams where they too distant from estuarine or marine habitats to may occur as far as 600 and perhaps 1000 kilo- and from which the young must travel to spend meters inland and at altitudes of as much as 925 at least a part of their early larval life in water meters. Almost all have been reported to live in with higher salinities. What distance is "too far" swiftly flowing water where they tend to be con- has not been determined. That flowing water is centrated on rocky substrates, although rooted not essential to these shrimps becomes evident in vegetation and debris appear in some places to view of the records for Atya intermedia in Crater provide adequate cover for numbers of them. Lake on Annobon and the pond on Dominica Many, if not most, of the females become oviger- mentioned by Fryer (see accounts below on A. ous, bearing several hundred eggs while surpris- intermedia and A. innocous). Although presumably ingly small (larger females produce upward to there was no current in the lake or pond, water 4000 eggs in a clutch)—the number of broods indeed flowed through them. produced by a single individual is not known. Possible exceptions to these generalizations ex- The eggs are retained beneath the abdomen of ist in the Nun and Osse rivers in Nigeria where the female until they hatch, and the escaping C.B. Powell (pers. comm.) found A. gabonensis in larvae, which live for some six days in the labo- the faster flowing sections of the rivers that are ratory (Hunte, 1979b), do not feed but must deep, lack rocks, and, during the dry season, are make their way to the mouth of the river where sluggish. in water of higher salinity they molt to the second instar. In A. innocous, a total of 12 stages (or instars) occurs over a period of 76 to 119 days (in Life History the laboratory) before the larvae metamorphose This account of the life history of the several into juveniles. Nothing is known of the habits of species of Atya is based upon field observations the larvae, but presumably the juveniles return 16 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 8.—Habitats exploited by members of the genus Atya, and implements used in Jamaica to catch freshwater shrimps: a, Mannet's Gutter, near mouth, in Dominica, frequented by Atya imocous; b, basket employed in Jamaica in catching both atyid and palaemonid shrimps; c, an upstream locality on Mannet's Gutter, Dominica, inhabited by Atya innocous and A. scabra; d, Layou River, Dominica; e, children using basket to catch "jongas" (freshwater shrimps) in small stream in St. Andrew Parish, Jamaica.

to the rivers,makin g their way upstream. As they shaded pools during daylight hours. Both juve- approach maturity, they become more active at niles and adults dig shallow depressions under night than during the day, although they are rocks or find existing crevices that satisfy their frequently seen moving over the substrate of need for cover. They feed either by filtering or by NUMBER 364 17 sweeping and/or scraping organic material from the substrate with the aid of the terminal tufts of setae on the first and second pereiopods. Even though the shrimp are capable of swimming, most of their locomotion is accomplished by walk- ing and climbing. As in many decapods, fast backward propulsion is accomplished by rapid flexion of the abdomen. Not only do they occupy riffle areas of rivers, but they also make their way up small tributaries almost to the source, and inasmuch as they traverse small cascades there is every reason to believe that in their migrations upstream they leave the water for at least short distances. It is not inconceivable, therefore, that they are able to cross low divides from one head- water stream to another. There are no data as to their longevity in the wild, but Fryer (1977:72) kept a specimen of A. innocous in the laboratory for about six years and nine months, and Abele (1975:56) reported that a specimen of A. dressleri Ancestor had lived in an aquarium for more than five FIGURE 9.—Dendrogram depicting relationships of members years. of the genus Atya.

Relationships and Dispersal Two species, A. lanipes and A. dressleri, exhibit FIGURE 9 these features, and there seems to be little reason In assessing the interrelationships of members to choose either as having preserved the larger of the genus Atya, we are in at least basic agree- number of primitive characters; however, the ment with Bouvier (1925). The more generalized combination of a slender, almost smooth merus species appear to us to be those in which (1) the of the third pereiopod and the tapered rostrum rostral margins tend to converge from base to causes us to favor A. lanipes as perhaps having apex, (2) the carapace is rather smooth, lacking more in common with members of other genera prominent ridges and spines, (3) the telson is of the family Atyidae. Were the rostral margins slender and elongate, (4) the ventral margin of of A. ortmannioides less suddenly contracted and the abdominal pleura lacks sclerotized denticles, the merus of the third pereiopod less tuberculate, (5) the basal segment of the antennule is devoid this shrimp would rival A. lanipes and A. dressleri of premarginal sclerotized denticles, (6) the coxa as being one of the most primitive members of of the third pereiopod has an entire, as opposed the genus. Whether a strong or weakly developed to scalloped, distal margin but lacks a lateral pterygostomian angle is the more generalized spine, corneous tubercles, and a caudomesial pro- seems problematical, but should a weak one be jection, (7) a comparatively slender merus on the the more primitive then at least in this respect A. third pereiopod, that is weakly tuberculate at dressleri merits consideration as representing the most, lacks more than a vestigial distal ventrome- ancestral Atya. sial spine that opposes a strong tubercle on the Closely allied to these three species are A. in- carpus, and (8) the dactyl is hinged and freely nocous, A. intermedia, and A. brachyrhinus. In the movable, and the flexor surface is provided with former two the rostral margins have become sub- two oblique (perhaps irregular) rows of tubercles. angular, and the merus of the third pereiopod is 18 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY often considerably more strongly developed and tral stock than has A. margaritacea is suggested by studded with tubercles, the apical parts of which the consistent presence of sclerotized premarginal are heavily cornified, and in many populations of denticles on the dorsal surface of the proximal A. innocous individuals have acquired sclerotized podomere of the antennule, by the contiguous or denticles on the ventral margin of the second subcontiguous linear series of squamiform tuber- through fifth abdominal pleura. The rostrum of cles on the flexor surface of the propodus of the A. brachyrhinus is very short, the acumen has be- same appendage, and by the presence of denticles come greatly reduced, and the rostral margins on the pleuron of the second abdominal segment. broadly rounded; moreover the section of the The most divergent species of the genus are cephalic margin of the carapace between the Atya crassa and A. gabonensis. Both display an array antennal and pterygostomian spines that is ex- of unique characteristics that distinguish them cavate in this shrimp is convex in A. innocous and from the more generalized members of Atya, and A. intermedia. Denticles are also present on the whereas they have several obvious advanced third through fifth abdominal pleura. Unfortu- characteristics, both have extended certain mod- nately only two females of A. brachyrhinus are ifications beyond that exhibited by the other. In available, and perhaps the merus in the illustra- superficial appearance, no doubt A. crassa seems tion of this specimen (Figure I5h) should not be to be the most divergent member of the genus; it compared with those in the other species that possesses a rostrum with convergent margins depict the third pereiopods of males. (probably secondarily acquired), but the dorsal Somewhat intermediate between the latter carina is uniquely provided with a row of spines, three species and the advanced Atya is A. africana, and most of the ridges (present also in A. gabonen- which has a rostrum with tapered margins and sis) adorning the carapace are studded with spines abdominal pleura that lack denticles on the ven- that are also unique in the genus. Whereas the tral surface. Premarginal denticles, however, are ocellar beak remains rather small and the tubercle present on the dorsal surface of the proximal on the sternum of the fifth abdominal segment podomere of the antennule; some of the subspi- moderately prominent, the ocellar beak of A. niform tubercles on the flexor surface of the pro- gabonensis is, by comparison, very long and blade- podus of the third pereiopod are arranged in a like, and the hornlike projection (enlarged tuber- contiguous or subcontinguous row, and the den- cle) on the sternum of the fifth abdominal seg- ticles on the flexor surface of the dactyl of that ment is decidedly more prominent than in other appendage are aligned in a single series. In all of members of the genus. Thus in different respects the aforementioned species, the tubercles of the both A. crassa and A. gabonensis seem to be the third pereiopod tend to be at least subspiniform. most divergent species assigned to the genus Atya. In those that follow, some, if not most, of the To subject our concept of the interrelationships tubercles of this appendage are distally flattened of the members of the genus Atya to a more and strongly cornified. objective test, we selected 25 traits (see list of Atya margaritacea and A. scabra are closely allied "Characters Compared") and, employing the species in which (1) the rostral margins are sub- Wagner Program of computer analysis (Farris, parallel or concave along the basal half and form 1970; Farris, Kluge, and Eckardt, 1970), con- a distinct angle, often produced, at the base of structed a series of cladograms, the most parsi- the acumen, (2) the punctations of the carapace monious of which is depicted in Figure 10. Com- bear short stiff setae, (3) the pleura of the second paring it with the more subjective (based upon or third through fifth abdominal segments bear "weighted characters") dendrogram in Figure 9, sclerotized denticles, and (4) the denticles on the the relative positions of Atya ortmannioides and A. flexor surface of the dactyl of the third pereiopod africana with respect to their close relatives seem are arranged in a linear series. That A. scabra has most in conflict. Data supporting the arrange- digressed slightly more from the supposed ances- ment in Figure 10 are evident in the cladogram NUMBER 364 19

•*• <*> oo 00 •». « a a > CO 1 •ao to •2. u a i I ! <*>

- brachyr • ortmam •S • innocow • margari - gabonen u < -o .2 2-4 3-2 9-2 4 — 2 11-2 C16-1) (•16-1) 7-2 17-2 (19-2) [2-3] 18-2 (23-2) 25—2 10 5-2 (23-2) (10-2) 8 12-3 13 — 2 15-2 (19-2) (12-2) 2 — 1 14-2 1-2 20-2 (8-2) 22-2 2-4 24 — 3 9 •2—2, *8—1 7

6 •12—2, 16—2, 24—2

4 (8-2)

3 21—2

2 2—3 * - Reversal 11 6 -2 ; - independent iicquisition

FIGURE 10.—Cladogram expressing relationships of members of the genus Atya. and list of characters considered. Reasons for the margins on the rostrum and of denticles on the groupings shown in Figure 9, however, are not so second through fifth abdominal pleura were con- concisely summarized. Perhaps the slender body sidered to be primitive retentions rather than and rather spindly ambulatory pereiopods (char- having resulted from character reversals. The acters that were not among those selected for arrangement of the tubercles on the flexor surface computer analysis) of A. ortmannioidcs led us toof the propodus of the third pereiopod is so much group this species with the supposedly primitive like that of A. scabra, yet possessing the form of A. dressleri and A. lanipes. The contrasting nature those in the less specialized A. innocous, that we of the same features, robust body and append- viewed those in A. africana as representing a tran- ages, together with others, suggested to us a recent sitional stage between the A. innocous-intermedia common ancestry of A. brachyrhinus with A. innocous stock and A. scabra. Moreover, the general mien and A. intermedia. of A. africana is so much more that of A. scabra and In considering the affinities of A. africana, we A. innocous than that of A. gabonensis and A. crassa were convinced that it shares more in common that one might question our having selected a with A. scabra and A. margaritacea than with other sufficient number of characters to be subjected to of its congeners. Both the absence of angular computer analysis. 20 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

CHARACTERS COMPARED As noted above, the postulated interrelation- ships of the members of the genus Atya as depicted (The first condition listed for each character is in Figure 9 are supported by morphological data considered to be plesiomorphic) involving several body regions, and there is every reason to believe that at least those species occur- 1. Carapace (1) glabrous, (2) hirsute ring on both the African and American continen- (2) tapering, (1) rounded, tal masses have existed in a comparatively stable 2. Rostrum (3) subangular, (4) angular state since at least early Cenozoic times, perhaps (1) spines absent, (2) spines since the middle Mesozoic. Inasmuch as the spe- 3. Median carina present cies comprising the genus are so similar in their (1) no spines, (2) spines 4. Carapace surface (1) ridges absent or weak, morphology, in such elements of their life histories 5. Carapace surface (2) ridges strong as are known, and in their habitat distribution as 6. Pterygostomian (1) weak, (2) moderate to strong well, one must conclude that they have had a angle monophyletic origin, one that predated a very 7. Second pleuron (1) denticles absent, (2) denticles wide separation of the African and American present continental masses. (Bouvier (1925:358) sug- 8. Third-fifth (1) denticles absent, (2) denticles pleuron present gested the influence of plate tectonics on atyid 9. Ocellar beak (1) short, (2) long and bladelike distribution.) The progenitors of modern Atya 10. Sternum, fifth (1) not hornlike, (2) hornlike with little doubt existed as recognizable members 11. Sternum, fifth (1) tubercle weak, (2) tubercle of the genus by the late Mesozoic (probably by strong early times). 12. Sternum, sixth (1) long, (2) intermediate, (3) short At least two distantly related species, Atya scabra 13. Preanal carina (1) compressed, (2) conical and A. gabonensis, had become differentiated and 14. Preanal carina (1) one spine, (2) two spines entrenched on the disjoined drifting African and 15. Telson (1) more than 1.5 times as long as American land masses by the advent of the Cen- wide, (2) less ozoic Era. Also, there is good reason to believe 16. Antennular (1) dorsal premarginal denticles peduncle I absent, (2) present that the progenitors of A. innocous and A. intermedia 17. Coxa, (1) ventrodistal margin not scal- (two species that differ in only minor respects) pereiopod III loped, (2) scalloped had likewise acquired the major features shared 18. Coxa, (1) anterolateral spine absent, by them, and segments of this common stock had pereiopod III (2) present also become established on the two continental 19. Coxa, (1) caudomesial projection small or pereiopod III absent, (2) strong blocks during or shortly after their juxtaposition. 20. Merus, (1) bowed, (2) straight The few differences that distinguish them could pereiopod III well have developed at any time during the Cen- 21. Merus, (1) lateral tubercles not strongly ozoic Era. The progenitors of Atya innocous are pereiopod III cornified, (2) strongly so likely also to have been ancestral to A. dressleri 22. Merus, (1) distoventral prominence absent and A. ortmannioides on the Pacific side of the pereiopod HI or not opposed; (2) strong and opposed by tubercle on carpus Middle American land mass and to A. lanipes and 23. Propodus, (1) flexor surface with tubercles A. brachyrhinus in the islands bordering the eastern pereiopod HI scattered or not contiguous in part of the Gulf of Mexico and Caribbean. mesial row, (2) those of As may be noted on the maps (Figures 19, 25, mesial row contiguous 28, 44) devoted to the distribution of Atya innocous, 24. Dactyl, (1) spinules on dactyl in two rows, pereiopod HI (2) single row, (3) cluster A. dressleri, and A. ortmannioides on the Pacific 25. Propodus- (1) movable, (2) fixed slope, A. innocous ranges from Nicaragua south- dactyl III ward to Panama; A. ortmannioides ranges from NUMBER 364 21

Baja California Sur southward to Acapulco, in latter, known only from a single cave on Barba- the state of Guerrero; and A. dressleri is known dos, may be a relict of a much more widely from only five localities in Panama. The distri- dispersed species (at least on that island), for bution of such closely allied species on the Pacific except for the limitation of dark pigment to the versant suggests that a part of the late Mesozoic eyes we have found no feature that suggests ad- wide-ranging pro-innocous stock reached the Pa- aptations to a spelean existence. (The reported cific slope and became isolated from that stock white body could well be due to contracted chro- present on land masses in or abutting the Gulf- matophores, a common response in pigmented Caribbean with the emergence of a complete to darkness.) Inasmuch as A. lanipes Middle American isthmus during the Eocene. has a broader range but is known to occur only That part of the stock moving northward ulti- in the Greater Antilles and Virgin Islands, it mately reached at least as far north as Baja seems reasonable to assume that the preservation California and gave rise to A. ortmannioides, of this remnant of the parent stock occurred in whereas that occurring in the more southern area the northeastern part of the Antillean region. became the immediate forerunner of A. dressleri. Supporting the suggested migrations of the What might have served later to separate these ancestors of modern A. innocous, A. dressleri, and A. two stocks was the submergence of part of the ortmannioides are the possible parallelisms that isthmus allowing the merger of waters from the could well have occurred in the continental Gulf-Caribbean and that of the Pacific—perhaps American origins and distributions of A. scabra across the Isthmus of Tehuantepec (note that this and A. margaritacea and of A. gabonensis and A. area lies between the ranges of A. ortmannioides on crassa. We suggest that accompanying the ances- the north and A. innocous and A. dressleri on the tral innocous stock (believed to be forebears of A. south). This event also allowed the innocous stock dressleri and A. ortmannioides) into the Pacific wa- from the Caribbean to gain access once again to tershed were members of both the ancestral scabra the Pacific watershed. With the last complete and gabonensis stocks that were the precursors of emergence of the Middle American isthmus, the A. margaritacea and A. crassa, respectively. With innocous stock on the Pacific became isolated from the emergence of the isthmus in the early Ceno- that on the Caribbean, and at the present time it zoic, these more advanced forms shared the seems to be less variable than populations occur- streams flowing into the Pacific with the ancestral ring in the West Indies. (This Pacific stock is that dressleri and ortmannioides, all four acquiring their designated Atya tenella by Smith, 1871.) On the distinctive features prior to the Eocene inunda- basis of our limited knowledge of the populations tion of parts of the isthmus. The crassa and mar- of this species on the Pacific slope, it seems to garitacea stocks ultimately ranged more widely have been more aggressive, or perhaps to have than did the other three {ortmannioides, dressleri, broader ecological tolerance, than does A. dressleri. and innocous), moving northward at least to the It is much more common and widespread in Tropic of Cancer, perhaps at times to much Panama where it may have largely displaced the higher latitudes, and in their southward migra- latter from much of its presumed former contin- tions almost reaching, or crossing the equator. uous range along the Pacific side of the Pana- Unlike many freshwater inhabitants, some, if not manian isthmus. all, members of the genus undergo their early There is little reason to choose one island or postembryonic development in saline waters and island group over another as representing the are thus probably able to gain access to river area in which divergence in the original wide- basins adjacent to those in which they hatched. spread pro-innocous stock led to the isolation of We are puzzled by the apparent absence of ancestors of A. lanipes and A. brachyrhinus. The Atya innocous and its relatives from the South 22 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

American continent, particularly in view of the characters: a rostrum with tapering margins (be- fact that it, the commonest species in the West lieved by us to be plesiomorphic (tying A. africana Indies and presumed by us to be the least eco- to the generalized A. lanipes and A. dressleri), along logically restricted of the American Atya, occurs with three features (denticles on the dorsal surface along both slopes of Panama and on Trinidad of the proximal podomeres of the antennule, a and Curasao. In contrast, the supposedly more row of contiguous spines on the flexor surface of ecologically limited A. scabra spans the range of the propodus of the third pereiopod, and a single the genus in the islands and continental slopes of row of denticles on the flexor surface of the dactyl the Caribbean and Gulf and ranges as far south of the same appendage) that link it with the more as the state of Santa Catarina in Brazil. advanced and sympatric A. scabra. In this in- In this attempt to correlate the distribution of stance, we are inclined to interpret the rostral the Middle American members of the genus Atya character as a secondary acquisition and to be- with events in the history of the isthmus joining lieve A. africana to have had its most recent com- the Middle and South American land masses we mon ancestry with the forebears of A. scabra. have neglected the occurrence of Atya scabra in a Because the known ranges of the two in Africa few localities on the Pacific slope. Perhaps mem- are almost identical and also that almost nothing bers of the species were introduced; if not, it is is known of their ecological distribution, little not inconceivable that they were able to cross the speculation concerning their past history on the continental divide, moving from a small head- continent seems warranted. water stream on the Caribbean slope to others emptying into the Pacific. Genus Atya (sensu stricto) As implied above, considerable stability must have been reached in the ancestors of Atya gabo- FIGURES 11, 12 nensis, A. scabra, and A. intermedia prior to the time Atys Leach, 1815:345 [type-species by monotypy: Atys scaber the American and African continents had become Leach, 1815:345; gender masculine; junior homonym of greatly disjoined. The fact that we are unable to Atys De Montfort, 1810:342 ()]. distinguish members of a Brazilian population of Atya Leach, 1816:421 [substitute name for Atys Leach, A. gabonensis from African representatives of the 1815:345; type-species by monotypy: Atys scaber Leach, 1815:345; gender feminine]. species, in combination with the improbability of Atia.—Latreille, 1817:37 [erroneous spelling]. intercontinental migrations or introductions, sup- Athys.—H. Milne Edwrds, 1838:352 [erroneous spelling]. ports such a conclusion. Virtually the same can Evatya Smith, 1871:95 [type-species by monotypy: Evatya be said of A. scabra, for whereas the absence of crassa Smith, 1871:95; gender feminine]. denticles from the second abdominal pleuron will Euatya.—Koelbel, 1884:317 [invalid emendation of Evatya Smith, 1871]. serve to distinguish the African representatives of AtyiaJ. Roux, 1932:564 [erroneous spelling]. the species from many, if not most, members of Atqya.—Chace and Hobbs, 1969:63 [erroneous spelling]. American populations, more than a few of the latter also lack such denticles. As is discussed Our concept of the genus Atya, similar to that below, A. intermedia is so closely allied to A. innocous of Chace (in prep.), is that it comprises the larger that we recognize its specific status with some members of the family Atyidae occurring in the reluctance. Thus these three shrimps have iden- tropics of West Africa and the Americas. Those tical or such close counterparts in the Americas species of the Indo-Pacific that formerly have that they must be considered little, if any, been assigned to this genus are excluded from it changed since the Africa-American continental by the restricted definition presented herein. In- masses were still approximate. asmuch as Chace has proposed their being as- An understanding of the affinities of A afncana signed to other genera prior to the inception of is clouded by a seemingly strange combination of our study, we refer the reader to his prospective NUMBER 364 23 publication for the generic dispositions of these The distribution of the species is summarized Indo-Pacific shrimps. in the accompanying tabulation. DIAGNOSIS.—Body pigmented, eyes well devel- oped. Rostrum not strongly compressed laterally, Africa median dorsal carina with or without spines, and Americas ventral keel with or without 1 to several teeth. Pacific Atlantic Atlantic Anterior margin of carapace armed with antennal Atya africana X and pterygostomian spines, latter sometimes re- Atya brachyrhinus X duced to angle; orbital spine lacking. Ventral Atya crassa X X margin of second through fifth abdominal pleura Atya dressleri X Atya gabonensis X X with or without sclerotized spinules. Telson with Atya innocous X X paired arched rows of 5 to 9 spines in adults; Atya intermedia X subacute to acute posterolateral angles exceeded Atya lanipes X by 2 pairs of mesially contiguous spines, more Atya margaritacea X mesial pair longer. Basal segment of antennule Atya ortmannioides X with or without dorsal corneous spinules proximal Atya scabra X X X to series bordering distal margin. Flagellar lobule of first maxilliped well developed. Third maxil- Conspicuously absent from this list of African liped never ending in spine, distal 2 podomeres and American shrimps assigned to the genus Atya with number of oblique rows of simple setae and is Atya serrata Bate (1888:699), which was de- with fewer strongly sclerotized, biserrate ones. scribed from Sao Antonio Valley, Sao Thiago, Pereiopods without exopods. First and second Cape Verde Islands. We have examined Bate's pereiopods with chelae completely divided, lack- syntypes and find that their affinities are with ing palm; fingers not gaping, subequal in length members of one of the Indo-Pacific genera rec- and tipped with brushes of long setae; carpus of ognized by Chace (in prep.). Only two reports of both appendages excavate distally, much shorter the occurrence of A. serrata within the range of than broad, and shorter than fingers. Branchial Atya (as defined herein) are known to us: the type- complement consisting of 5 pleurobranchs, 3 ar- locality and Muhlenburg Mission, Liberia (Bou- throbranchs, 1 podobranch, 5 epipods, and 5 vier, 1925:295). Subsequent collections made in mastigobranchs. First pleopod of male with en- the Cape Verde Islands have not contained rep- dopod broadly ovate, it preaxial surface armed resentatives of this species, and the specimen upon with many curved spines; appendix interna aris- which Bouvier reported its occurrence in Liberia ing from mesial surface of endopod; first pleopod has been examined by us and found to be a small of female with endopod tapering and lacking male (carapace length 7.1 mm) that clearly be- appendix interna. Second pleopod of male with longs to the African-American genus Atya and is appendix masculina compressed, slipper-shaped, almost certainly a member of Atya africana. In flattened mesially, spines confined to margins and view of these observations, and because of the mesial face. apparent affinities of the types with Indo-Pacific RANGE.—The members of the genus Atya are atyids, there is good reason to believe that Bate's confined to the American and African continents. specimens were not collected in the Cape Verde On the former it ranges on the Pacific versant Islands or elsewhere in the Atlantic versant of from 28°N to about 8°S, whereas in the Atlantic either Africa or the Americas. Holthuis (1951:26) watershed it occurs from the Tropic of Cancer to expressed a similar opinion in stating: "It is about 27°S. In Africa, the range is confined to doubtful whether this species, which is widely the Atlantic slope and is situated between 15°N distributed throughout the Indo-west pacific area, and 9°S. also occurs in West Africa." 24 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 11.—Lateral views of members of genus Atya (numbers in parentheses = carapace length in mm): a, A. africana, Mount Coffee, Liberia (6*, 22.6); b, neotype of A. scabra, Misantla, Mexico ((5, 29.8); c, A. scabra, Sao Tiago, Cape Verde Islands (6\ 24.5); d, syntype of A. margaritacea (6, 24.5); e, A. margaritacea, Mt. Chiriqui, Panama (8, 25.8);/, A. gabonensis, Volta River, Ghana (6\ 28.2); g, A. aassa, Rio Dagua, Colombia (6\ 37.8). NUMBER 364 25

FIGURE 12.—Lateral views of members of genus Alya (numbers in parentheses — carapace length in mm): a, A. irmocous, Mannet's Gutter, Dominica (<$, 24.4); b, syntype of A. tenella (= A. irmocous) (<$, 16.2); c, holotype of A. ortmanmoides (9, 15.9); d, A. intermedia, effluent of Crater Lake, Annobon (6*, 25.4); e, A. lanipes, Rio Maricoa, Puerto Rico (<5, 21.0);/, holotype of A. dressleri (6, 20.4); g, holotype of A. brachyrhinus (?, 15.7). Key to Species of the Genus Atya 1. Cephalic section of carapace with median dorsal row of spines ex- tending onto rostrum; prominent ridges on carapace also studded with spines A. crassa 1'. Cephalic section of carapace and rostrum lacking median dorsal row of spines; ridges present or absent from carapace, if present never studded with spines 2 2(1')• Rostral margins with angular (acute or right angle, never obtuse) bend or lateral prominences; tubercles on flexor surface of dactyl of third pereiopod arranged in small cluster or in single row ... 3 2'. Rostral margins with or without subangular bend, if present obtuse or with extremities rounded; tubercles on flexor surface of dactyl of third pereiopod never arranged in cluster or single row except latter in A. africana in which rostral margins tapering from base 5 3(2). Flexor surface of propodus of third pereiopod with some of heavily cornified squamous tubercles arranged in longitudinal rows, at least part of those in mesial row quite or nearly contiguous; most American representatives with row of sclerotized denticles on ven- tral margin of second abdominal pleuron A. sea bra 3'. Flexor surface of propodus of third pereiopod with few to many heavily sclerotized elevated or squamous tubercles, if some forming linear series, those in mesial row never nearly or quite contiguous; sclerotized denticles never present on second abdominal pleuron 4 4(3'). Carapace strongly sculptured; flexor surface of propodus of third pereiopod with conspicuous tufts of long plumose setae and with very few squamous tubercles; corresponding surface of dactyl of same appendage with single or cluster of tubercles just proxinnal to sclerotized tip; sternum of fifth abdominal segment with median curved hornlike projection A. gabonensis 4'. Carapace comparatively weakly sculptured; flexor surface of propo- dus of third pereiopod without conspicuous tufts of long plumose setae, with prominent squamous tubercles; corresponding surface of dactyl of same appendage with single row of tubercles; sternum of fifth abdominal segment with small median tubercle A. margaritacea 5(2'). Dorsal surface of proximal article of antennular peduncle with 1 to 3 sclerotized spinules; flexor surface of propodus of third pereiopod with tubercles arranged in 2 subparallel rows, those in mesial row subcontiguous and none between rows; corresponding surface of dactyl of same appendage with tubercles arranged in single row A. africana 5'. Dorsal surface of proximal article of antennular peduncle lacking sclerotized spinules; flexor surface of propodus of third pereiopod with scattered tubercles, if rows evident always with other tubercles between them; corresponding surface of dactyl of same appendage disposed in 2 rows, more proximal one almost always situated proximomesial to distal row 6 6(5'). Rostrum with margins strongly convex laterally, apex falling far short of second segment of antennular peduncle A. brachyrhinus 6'. Rostrum with margins subparallel to strongly convergent, apex reach* ing or almost reaching base of second article of antennular peduncle 7 7(6'). Rostral margins with strong subangular bend; ventral margin of third through fifth abdominal pleura with or without sclerotized spinules 8 7'. Rostral margins tapering from base; ventral margin of third through fifth abdominal pleura without sclerotized spinules 10 8(7). Lateral surface of merus of third pereiopod lacking conspicuous tubercles, all small and weakly, if at all, cornified; cephalolateral surface of carapace straight; rostrum, with acumen longer than basal part, directed anteriorly A. ortmanakoides 8'. Lateral surface of merus of third pereiopod studded with tubercles; cephalolateral surface of carapace convex (anterior to posterior); rostrum directed anteroventrally 9 9(8'). Ventral margin of third through fifth abdominal pleura, except in specimens from Pacific Basin, usually with sclerotized spinules; median tubercle on sternum of fifth abdominal segment compara- tively inconspicuous; preanal carina with single spine A. ianocous 9'. Ventral margin of third through fifth abdominal pleura always devoid of sclerotized spinules; median tubercle on sternum of fifth abdom- inal segment comparatively large; preanal carina with 2 spines A. intermedia 10(7'). Pterygostomian angle produced in prominent spine; dorsal surface of penultimate podomere of antennular peduncle with scattered spines; merus of third pereiopod without tubercles dorsally or laterally; preanal carina with vestigial spine; dorsal surface of telson with rows of 5 to 7 spines A. ianipes 10'. Pterygostomian angle weak or obsolete; dorsal surface of penultimate podomere of antennular peduncle with lateral row of about 6 spines and 1 to 3 more mesially situated; merus of third pereiopod with distinct tubercles dorsally and laterally; preanal carina produced in well developed caudally directed spine; dorsal surface of telson with rows of 7 to 9 spines A. dressleri

Arya africana Bouvier Atya senata.—Bouvier, 1925:295 [in part].—Holthuis, 1951:25, 26 [in part]. FIGURES 1A, 9-1 la, 13, 14 Atya gabonensis.—Holthuis, 1951:25 [in part]. Atya scabra,—Rathbun, 1900:313 [in part]; 1901:119 [in REVIEW OF LITERATURE.—The first reference to

part].-De Man, 1925:28 [in part]. this shrimp was that of Rathbun (1900) who africana Bouvier, 1904:138 [type-locality: "Samkitta _^ , V .. I-T-U- ^ •ynwiH Duuvra, ,7«V "^ WXATD ROO I*I reported the occurrence of rAtya scabra in Liberia, dans la riviere Ogooue," Gabon; type: MHNP 593, 1

(1925). A single male from the Ogooue River at from off Cameroon. He suggested (p. 23) that the Samkitta, Gabon, constituted the material on differences noted which Bouvier (1904) briefly described Atya afri- may be due to the different stages of development. The cana, largely contrasting it with his concurrently specimens from Boma show more resemblances to the adult described A. intermedia. The features pointed out form (absence of exopods on the pereiopods, shape of the were the longer, narrower rostrum on which the telson), than the specimens from Stations 57 [off Liberia] dorsal carina is not deflected apically, the strong, and 119 [off Cameroon]. The smallest specimen from Boma, toothed ventral rostral carina, and the well-de- 9 mm. in length, is, however, quite identical with the other veloped armature of the merus of the third pe- material from the same lot and is different from the largest specimen from Sta. 119. The Boma specimens were collected reiopod. Bouvier (1905) illustrated the cephalic in fresh or brackish water, while the specimens from Sta. 57 region of the holotype in dorsal and lateral aspects and 119 were taken in pure sea water. This might indicate and amplified his earlier description by adding that the early stages of this species are spent in the sea and that the third legs are enormously developed and that, when growing older, it returns to fresh water, just as is covered with numerous irregularly dispersed cor- the case in some species of the Palaemonid genus Macrobrach- neous tubercles, and furthermore that the inferior border of the merus of this appendage possesses Holthuis' (1951:25, 26) reference to Bouvier's a movable spine. In the account of this species in (1925) report of A. serrata from Muhlenberg Mis- his monograph of the family, Bouvier (1925) sion (see above) was based upon Bouvier's misi- contrasted A. africana with Atya moluccensis De dentification of Atya africana. Lebour (1959) cited Haan (1850:186), describing and illustrating the a new Liberian record for the species. former much more completely. A number of the In describing Atya lanipes, Holthuis (1963) characters chosen, however, are shared by other stated that A. africana was its closest relative and members of the genus, and few of those men- that the former could be distinguished from the tioned are useful in distinguishing A. africana from latter by the more slender third pereiopod of the its congeners. Only three specimens of A. africana adult male, the shape of the endopod of the first were reported by Bouvier, the holotype and a pleopod, and by the appendix masculina. Follow- male and female from Booue, Gabon. A fourth ing his redescription of Atya intermedia, Holthuis specimen was available to him, for the single (1966) noted its close relationship to A. africana shrimp in the collection of the (former) United and corroborated the existence of the differences States National Museum from the "Muhlenberg between the two species pointed out by Bouvier Mission, W. Africa, OJ. Cook" referred to by (1925). Vogel and Crew (1965) recorded this Bouvier (1925:295) and assigned to Atya serrata shrimp from the Mungo River, Cameroon. Bate (1888), is a member of A. africana. (See locality 4 in Liberia below). Monod (1967) reported Atya africana to be a large species (as long as 11 cm) occurring in De Man (1925) recorded the presence of this Gabon and Congo; the illustrations were those of shrimp in Mbuma, Zaire, and Holthuis (1951) Bouvier (1925) and De Man (1925). Disney tentatively assigned juveniles collected from off (1969) and Lewis, Disney, and Crosskey (1969) Liberia and Cameroon and from Boma, Zaire, to found that the young stages of certain phoretic this species. Holthuis described the specimens blackflies utilize Atya africana as a host in the Bille, from the latter in considerable detail and pointed Meme, and Mungo rivers in West Cameroon. In out several striking differences between these ju- much expanded reports, Disney (1971, 1975) pre- veniles collected in fresh water and those from sented his observations on this shrimp in tribu- the two marine localities. He illustrated the ce- taries of the Mungo and Meme rivers near phalic region, caudal part of the telson, and Kumba, West Cameroon, where it is infested with several appendages of the specimens from Boma the larvae and pupae of three species of blackflies and the caudal part of the telson of one of those (Simuliwn dukei, S. kenyae, and S. damnosum). Most NUMBER 364 29 of the adult shrimp were found under stones, mandible, first maxilla, distal podomeres of the whereas the smaller individuals were more abun- first, third, and fifth pereiopods, and caudal part dant among tangles of leaves. The composition of the telson. These drawings were based on of a population as revealed by netted specimens specimens from off Cameroon and from Boma. was roughly 57% having a total carapace length DIAGNOSIS.—Cephalic region of carapace not (c.l.) of 6 to 15 mm, about 23% with 16 to 20 mm conspicuously sculptured, lacking spines other c.l., 11% with 21 to 25 mm c.l., 6% with 26 to 30 than antennal, pterygostomian, and ventral ros- mm c.l., and the remaining 3% with 31 to 45 mm tral; pterygostomian spine strong. Rostrum with c.l. He found that in 1968-1969 the ovigerous margins tapering from base, lacking angular females, which possessed total carapace lengths bends. Ventral margin of all abdominal pleura from 11 to 25 mm, occurred in greatest numbers usually lacking sclerotized spinules (occasionally from October to November and from April to present on fifth), and caudoventral angles of May (ratio of 8:5) with an apparent "breeding fourth and fifth pleura not produced in spines. pause" from December to March. The Simulium Sternum of fifth abdominal segment with small larvae occur in the gill chamber or among the median tubercle, that of sixth about 1.5 times as cephalothoracic appendages of the shrimp, and broad as long. Preanal carina with compressed the pupae were found only on the ventral side of spine not overreaching posterior extremity of the scaphocerite. basal part of carina. Telson 1.8 to 1.9 times as Crosnier (1971) found juvenile specimens, long as broad with 4 or 5 spines in each of 2 rows. which he tentatively assigned to A. africana, at a Antennular peduncle with dorsal surface of prox- depth of about 10 meters in the Bay of Pointe imal article bearing 1 to 7 sclerotized spinules Noire, Congo. He stated that they resemble the proximal to distal row; penultimate article 1.1 to specimens from off Cameroon described by Hol- 1.5 times as long as wide and dorsal surface with thuis (1951) and not those from the freshwater several to many spinules, some of which some- habitat in the Belgian Congo (= Zaire). Lemasson times arranged in sublateral row. Coxae of third (1973) added no new information but expressed and fourth pereiopods lacking prominent antero- the opinion that this shrimp, which attains a total lateral spine. Third pereiopod with merus length of 11 cm, might lend itself to cultivation rounded ventrally, 2.9 to 4.1 times as long as in Africa. high, ventromesial surface slightly to strikingly PUBLISHED ILLUSTRATIONS.—The following il- bowed, never parallel to that of corresponding lustrations of Bouvier's Atya africana are available. podomere of other third pereiopod, and lateral Dorsal and lateral views of the cephalic extremity surface rather strongly tuberculate, tubercles sub- of the carapace were presented by Bouvier (1905) linearly arranged; propodus 2.1 to 3.2 times as and republished by him (1925) along with others long as wide, extensor surface studded with of the first pleopod of the male, the appendix strongly sclerotized subspiniform tubercles, and masculina, the caudal angles of the fourth and flexor surface with similarly sclerotized scalelike fifth abdominal segments, the uropodal angle, to spiniform ones somewhat linearly arranged, and the caudal region of the telson; in the same those comprising row immediately mesial to me- year, dorsal and lateral views of the cephalic dian line contiguous to subcontiguous; dactyl region were depicted by De Man (1925). Monod freely movable and bearing single row of corneous (1967) reproduced the illustrations of Bouvier scalelike to spiniform denticles on flexor surface. (1925) along with those of De Man (1925). MALE (Cavalla River, Bolobo, Liberia).—Ros- Illustrations of juvenile specimens tentatively trum (Figure 13a, d) with margins tapering from assigned to this species by Holthuis (1951) include base, bearing only slightest angle delimiting pos- dorsal and lateral views of the cephalic region as terior end of acumen; apex of latter reaching no well as drawings of the basal antennular segment, more than 0.25 length of penultimate podomere 30 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 13.—Atya afncana (all from male from Cavalla River, Bolobo, Liberia, except eji,i from Mt. Coffee, Liberia): a, dorsal view of cephalic region; b, dorsal view of antennular peduncle; c, dorsal view of telson; d, lateral view of cephalic region; e, flexor surface of distal part of fourth pereiopod;/, lateral view of preanal carina; g, lateral view of second through fifth abdominal pleura; hjc, lateral view of third pereiopod; i, flexor surface of distal part of Fifth pereiopod; j, mesial view of appendices masculina and interna, /, flexor surface of distal part of third pereiopod. (Scales marked in 1 mm increments.) NUMBER 364 31 of antennule; dorsal median carina gently curved, proximal to lateral row. Antenna with ventrolat- not excavate dorsally, not dipping below level of eral spine on basis reaching about same level as lateral carinae posterior to acumen, and reaching left stylocerite but not quite so far as right (latter apex of acumen; ventral carina with single weak attaining base of distal third of proximal segment tooth at base of apical fourth of rostrum; ocellar of antennular peduncle); lateral spine on scapho- beak, hidden in lateral view by eyes, with dorsal cerite strong, extending much beyond tip of ros- margin only slightly arched and meeting vertical trum, almost to end of antennular peducle; la- anterior margin at right angle, its anterior ex- mella overreaching latter; flagellum extending to tremity not approaching level of tip of stylocerite. fifth abdominal tergum. Antennal and pterygostomian spines moderately Third maxilliped overreaching antennular pe- strong and acute; no spines present between duncle by almost half length of distal podomere them. Surface of carapace with many small, shal- of endopod; tip of exopod reaching base of distal low, setiferous punctations, but devoid of ridges third of penultimate podomere of endopod; latter and spines other than those just mentioned; setae between 1.3 and 1.4 times length of ultimate borne in ventral submarginal punctations not podomere. conspicuous. First pereiopod reaching distal end of anten- Pleura of first three abdominal segments (Fig- nular peduncle, second reaching base of distal ure I3g) with rounded posteroventral extremities; fifth of fingers of first pereiopod; terminal brush corresponding parts of fourth and fifth angular of setae of both appendages lacking scraping but not produced in spines. All pleura lacking denticles. Third pereiopod (Figure I3k,l) with corneous spinules on ventral margin, but fifth lateral distoventral spine on merus and carpus, with conspicuous fringe of plumose setae. Fourth and, when extended anteriorly, overreaching an- abdominal tergum about 1.1 times as long as fifth tennular peduncle by dactyl and four-fifths (sixth subequal in length to fifth) and subequal length of propodus; merus with ventromesial in length to telson; sixth tergum only slightly margin bowed, about 4 times as long as high, 2.2 shorter than telson. Sternum of fifth abdominal times as long as carpus, and 1.8 times as long as segment with small laterally compressed median propodus; latter 2.7 times as long as wide, 1.2 tubercle (Figure Ih). Sternum of sixth segment times as long as carpus; distoventral margin of almost 0.7 as long as broad. Free part of preanal coxa entire (weakly undulate), and mesial cau- carina (Figure 13/) very short, apex not reaching doventral prominence lacking conspicuous setal level of caudal extremity of basal part of sclerite. clusters. Lateral, dorsal, and ventral surfaces of Telson (Figure 13c) little more than twice as long merus studded with many small apically sclero- as broad, its dorsal surface bearing paired concave tized tubercles, most arranged in sublinear series; rows of 5 corneous denticles and posteromedian clusters of plumose setae flanking tubercles, most tubercle, latter slightly overhanging caudal mar- conspicuous tufts forming oblique row on lateral gin. surface; mesial extremity of podomere weakly Proximal podomere of antennule (Figure 13b) produced at level of mesial articular condyle of with stylocerite reaching base of distal third or carpus. Carpus strongly tuberculate except ven- fourth of segment; dorsal surface with linear clus- trally where fewer tubercles, but clusters of long ter of setae and 2 corneous spinules; distal margin plumose setae present, especially ventrolaterally. bearing row of 7 (right) or 9 (left) corneous Propodus with most tubercles arranged in linear spinules; penultimate segment of peduncle about series, 2 parallel rows on flexor surface well de- 1.2 times as long as wide and bearing 20 spinules fined. Dactyl movable, its flexor surface with on dorsal surface and 10 on distal margin; ulti- single longtitudinal row of 4 denticles flanked mate podomere with row of 6 spinules at base of distally by paired clusters of setae. lateral flagellum, 3 at base of mesial one, and 4 Fourth pereiopod with dactyl reaching end of 32 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY proximal third of propodus of third pereiopod; Coffee, St. Paul River, 16* (22.9), 21 Apr 1897, length of merus 2.5 times that of carpus, and R.P. Currie. (4) fUSNM, Muhlenburg Mission, latter 0.7 as long as propodus. Fifth pereiopod 16* (7.1), May 1892, OFC. (5) *BM, St. John reaching articulation of carpus and propodus of River near Yila, at 190m, 16* (27.7), 2 Mar 1971, fourth; merus about 1.3 times as long as carpus, RG. (6) fUSNM, Cavalla River at Bolobo, 56* latter approximately 0.8 as long as propodus. (11.4-18.6), 2 ovig? (10.8-13.9), Jun 1946, H.A. Ornamentation of merus, carpus, and propodus Beatty. (7) fBM, Cavalla River near Nyaake of fourth pereiopod similar to that of third except (4°52'N, 7°35'W), Grand Gedeh Co, 16* (35.2), for additional spine on ventral surface of merus 13 Dec 1970, RG. (8) fBM, Monot River at base of distal third, 3 distolateral spines and (4°51/N, 7°39;W), Grand Gedeh Co, 16* (14.3), conspicuous row of long plumose setae on lateral 16 Oct 1969, RG. (9) fBM, Masowa Creek at surface of carpus and propodus. Ornamentation Bendaja (7°9/N, 11°15'W), Grand Cape Mount of fifth pereiopod like fourth but with 2 ventral Co, 1? (9.1), 14 Apr 1971, RG. (10) fBM, Ma- spines on merus proximal to distal ventrolateral sowa Creek near Dambala, Grand Cape Mount spine, and 3 distolateral spines. Co, 46* (6.1-13.3), 1$ (8.5), 14 Apr 1971, RG. (11) Diaresis of lateral ramus of uropod flanked fBM, Koejar River near Mano River Mine proximally by row of 17 articulated corneous (7°16'N, ll°10'W), 16* (19.2), 1 ovig? (13.8), 18 denticles and slightly larger fixed spine at lateral Jun 1969, RG. (12) fBM, Mafa River near Ben- end of row. douma (7°12'N, 10°59'W), Grand Cape Mount COLOR NOTES.—According to C.B. Powell Co, 1$ (8.6), 17 Jun 1967, RG. (13) fBM, Yah (pers. comra.), in Nigeria this shrimp is rich River near Zangbuck Town (6°50'N, 9°6'W) at brown with a yellowish to tan mid-dorsal stripe, 200 m, 1$ (17.9), 2 Mar 1971, RG. (14) fBM, bordered with black, extending from the rostrum Lower Given River (6°8'N, 9°4'W), Nemba Co, almost to the tip of the telson, and there is a dark 39 (9.1-13.0), 4 May 1971, RG. (15) fBM, Koene band across the tail fan in young individuals. near Bestmantown, Kulu, Sinse Co, at 100 m, SIZE.—The largest specimen of which we are 136* (6.5-23.7), 12? (6.9-17.3), 2 ovig 9 (12.9, aware is a male with a carapace length of 35.2 14.7), 19 Mar 1970, RG. (16) Atlantide sta 57 mm. The largest female measured by us has a "off Liberia" (5°59'N, 10°27/VV), bottom sample, carapace length of 21.0 mm. The smallest and 1 juv (9.0, t.l.) questionably assigned to this spe- largest ovigerous females have corresponding cies, 8 Jan 1946 (Holthuis, 1951:20). (17) Atlan- lengths of 11.1 and 20.4 mm, respectively. tide sta 53 off Port Marshall (Lebour, 1959), "a DISTRIBUTION AND SPECIMENS EXAMINED.—Atya few specimens" questionably assigned to this spe- africana ranges from Liberia southward to Zaire cies, 4-7 Jan 1946. (Figure 14). Records for the known localities are NIGERIA: (1) fBM, Ataiyo River at Oban Hills, listed below. Collections that we have examined at 150 m, 1? (12.6), P.A. Talbot. (2) Affluent of are marked with an asterisk if they have been Akpe-Yafe River at Aking, Cross River State previously reported and with a dagger if they are (5°25'N, 8°37'E) (fide C.B. Powell). (3) near reported herein for the firsttime . Numbers follow- Ogoja, migrating juveniles (fide CBP). (4) Nun ing the specimens listed are measurements, in River at Kaiama, migrating juveniles (fide, mm, of the carapace length or, if followed by CBP); 17 juv (3.7-5.2), 22 Jun 1977, CBP (these "t.L," total length. Some listings lack the date the specimens from the Nun River are tentatively collection was made and/or the name of the assigned to this species). collector; these could not be determined. CAMEROON: (1) *BM, Blackwater River LIBERIA: (1) fUSNM, Beulah, Bassa Co, 1$ (4°22'N, 9°47'E) (Lewis, Disney, and Crosskey, (12.9), Feb 1896, O.F. Cook. (2) fBM, St. Paul . 1969:232), 26* (10.6, 20.0), 1 juv (5.8), R.L.H. River near Hendi, Bong Co, 16* (8.2), 1 ovig ? Disney; 16* (25.6), 7 Oct 1968, RLHD; 26* (14.5, (13.5), 13 Jun 1969, R. Garms. (3) fUSNM, Mt. 24.8), 2? (14.2, 20.4), 1 juv (7.0), 3 Dec 1968, NUMBER 364 33

FIGURE 14.—Distribution ofAtya afncana.

RLHD; 26 (27.1, 28.5), 11 Dec 1968, RLHD; 26 (8.1-21.0), 1 ovig 9 (12.2), 12 Apr 1969, RLHD; (28.0, 31.8), 13 Feb 1969, RLHD; 66 (8.9-20.1), 1 ovig 9 (15.8), 28 Apr 1969, RLHD; 27 juv 6$ (8.5-17.7), 2 ovig 9 (13.1, 18.0), 1 juv (5.3), 17 (4.9-6.2), 23 Oct 1970, RLHD. (4) fBM, Kobe Mar 1969, RLHD; 46 (9.8-16.0), 6$ (8.8-16.9), Let, 1 juv (4.6), 11 Jun 1970, RLHD. (5) fBM, 22 Mar 1969, RLHD; 1<5 (18.8), 1$ (12.1), 1 ovig Kobe River near Etam, 1 ovig 9 (20.0), 6 Dec 9 (19.6), 31 Mar 1969, RLHD; 5

(12.1), 29 Oct 1969, RLHD. (8) *BM, Mungo antennular peduncle and exhibits as many as River near Boduma (Vogel and Crewe, 1965), 1 three ventral spines, but they may be lacking. ovig 9 (14.8), 8 juv (5.1-6.5), 26 Nov 1968, The range of variation in the number and dis- RLHD. (9) Atlantide sta 119 (2°55'N, 9°21'E), position of the corneous spinules on the anten- stramin net 100 cm, 3 juv 6* (2.3-3.0) (question- nular peduncle is mentioned in the "Diagnosis." ably assigned to this species; Holthuis, 1951:20), None of the variations noted seems to be restricted 28 Feb 1946. to a limited part of the range of the species. GABON: (1) Booue, 16*, 1$ (Bouvier, 1925:307). Like those specimens cited above that were (2) *MHNP, Ogooue River at Samkitta (Bouvier, questionably assigned to this species by Holthuis 1904:138), 16* (46.0, holotype), Mar 1877. (3) (1951) and Crosnier (1971), the juvenile female fMHNP, Ogooue, 16 (14.3), M. Marche. listed from the third locality in Zaire is only CONGO: "baie de Pointe-Noire, 10 m env.," 54 tentatively assigned to this species. The absence juv (9.9-11.5, t.l.), A. Crosnier (Crosnier, 1971: of a pterygostomian angle and only three spines 570, questionably assigned to this species). in each of the two rows on the dorsal surface of ZAIRE: (1) MBuma dans le Mayumbe (De Man, the otherwise typically appearing telson are 1925:26), 16* (96, t.l.), 2 ovig 9 (65, 70, t.l.), 7 unique among the specimens we have examined. other specimens, 25-26 Oct 1920, Dr. Schoute- The third pereiopods are so poorly developed that den. (2) *BM, Atlantide sta 127, Boma, 4 juv they are not helpful in identifying the specimen. (2.3-4.2), (questionably assigned to this species; The long tapered rostrum is the single character Holthuis, 1951:20), 10 Mar 1946. (3) ? fMCZ, that suggests its being a young member of Atya Zaire River near Bulu, W of Luozi (5°1'S, africana. 14°1'E) swimming at surface in protected cove, ECOLOGICAL NOTES.—There is little informa- 1 juv 9 (7.1), 15 Jul 1973, T.R. Roberts, DJ. tion concerning the habitats exploited by mem- Stewart (see "Variations"). bers of this species except that these shrimp fre- VARIATIONS.—Few variations worthy of men- quent creeks and rivers from sea level to 200 m. tion have been noted. Most appear to be associ- A juvenile with a carapace length of 7.1 mm, ated with the size, and presumably the age, of the believed to be a member of this species, was individual. This is particularly true of the larger caught swimming at the surface of a protected spines on the various podomeres of the ambula- cove in the Zaire River near Bulu, Zaire. If the tory pereiopods; the smaller specimens, for the juveniles described by Holthuis (1951:20-23) and most part, have more, a full complement consist- Crosnier (1971:571) prove to be members of A. ing of a series of three spines on the ventrolateral africana, then a part of the life cycle is probably surface of both the merus and carpus of all three typically spent in salt water at depths ranging legs; the spines are progressively larger from prox- from five to 62 meters; at least the young can imal to distal end of each podomere, and the tolerate sea water within this depth range. As most persistent with age of the shrimp is the distal pointed out by Holthuis (pp. 20, 23), juveniles ventrolateral member on the merus. The fringe that may be conspecific and that are smaller than of setae that extends along the lateral surface of the largest found in salt water also occur in fresh the merus distally to about midlength of the water. propodus of the same pereiopods is also best Disney (1971), working in streams in the vicin- developed in comparatively small individuals ity of Kumba, West Cameroon, found that most that have recently molted, but in the more crusty, of the adult members of the A. africana population larger males, little more than a trace of the row occurred under stones and that the smaller indi- is evident on the merus, and it is often reduced viduals were found more frequently among leaves throughout its length. The rostrum is variable in and other debris, suggesting to him a difference length, attaining the basal one-third to almost of preference in the microhabitats of the smallest two-thirds of the penultimate podomere of the and adult members of the species. A brief account NUMBER 364 35 of the population structure reported by him is most of its features it resembles Atya dressleri, A. presented in the "Review of Literature" above. lanipes, and, to a lesser extent, A. innocous and A. We were informed by C.B. Powell (pers. intermedia. The acuminate rostrum, the rather comm.) that his specimens taken at Aking, Ni- weakly developed tubercles on the merus of the geria, came from a wide, shallow river with a third pereiopod, the proportions of this podomere, bottom made up almost entirely of rocks and and the absence of corneous spinules on the ven- boulders, and where there was a swift current. tral margin of the abdominal pleura align this The specimens were caught among accumula- shrimp with A. dressleri and A. lanipes. The mod- tions of dead leaves and other debris wedged erately short sternum of the sixth abdominal between rocks in places where the current was segment is more like that of A. dressleri, A. innocous, fastest. and A. intermedia than like that of A. lanipes, but LIFE HISTORY NOTES.—Ovigerous females have the preanal spine is quite short. Certainly it ap- been obtained from March to June and from pears to be more closely allied to the four just October to December. To our knowledge, collec- mentioned species than it is to A. scabra and A. tions have not been made during other months of margaritacea. Nevertheless the ornamentation of the year. Reference to the juveniles in "Ecological the flexor surface of the propodus and ischium of Notes" above suggests that if those in question this appendage is positioned as it is in A. scabra, are members of this species, like the larval mem- that is, the spines on the mesioventral surface of bers of other species of the genus studied by the propodus are arranged in a row, and are Hunte (1975, 1977, 1979b), the first larva must subcontiguous, and those of the dactyl are dis- return to salt or brackish water. At least some of persed in a single series. Individually, however, these juveniles occur in such habitats when they the spiniform tubercles of the entire appendage, attain a total length of 12 mm (Holthuis, especially those on the flexor surface of the pro- 1951:20). De Man (1925) reported the eggs of podus, resemble those of A. dressleri and A. lanipes ovigerous females from "MBuma," Zaire, to be more closely than they do the more heavily cor- 0.32 mm wide and 0.58 to 0.62 mm long. In nified discoid ones in A. scabra and A. margaritacea. specimens from the Blackwater River, Cameroon, Cornified spinules on the dorsal surface of the we found eggs ranging from 0.3 to 0.4 mm wide basal segment of the antennule are similar to by 0.6 to 0.7 mm long. those in A. scabra. Disney (1971) has added more than any other to our knowledge of the life history of this shrimp. Atya brachyrhinus, new species In the vicinity of Kumba, West Cameroon, he found that the ovigerous females, ranging in total FIGURES U, 9, 10, I2g, 15 carapace length from 11 to 25 mm (about half of them 16 to 20 mm) appeared in greatest numbers DIAGNOSIS.—Cephalic region of carapace not from October to November and about five- conspicuously sculptured, densely covered with eighths as many from April to May. An apparent fine setae borne in small punctations, lacking "breeding pause" occurs from December to spines other than antennal and pterygostomian; March. Juveniles of small size, which were be- rostrum with convex margins tapering anteriorly, lieved to be migrating, were collected by an as- only slight irregularity in position corresponding sistant to C. B. Powell (pers. comm.) near Ogoja to angle in some congeners; apex not nearly and Kaiama, Nigeria. reaching distal extremity of proximal segment of REMARKS.—Because of the tapering rostral antennule; pterygostomian angle acute but dorsal margins of this shrimp, its identity is as readily margin not excavate. Ventral margin of third, recognized among the African members of the fourth, and fifth abdominal pleura with linear genus as is the highly sculptured Atya gabonensis. clusters of sclerotized spinules; caudoventral an- Its affinities, however, are not entirely clear. In gles of fourth and fifth pleura not produced in 36 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 15.—Atya brachyrhimts (all from holotypic female except b, i,j from paratypic female): a, dorsal view of cephalic region; b, dorsal view of antennular peduncle; c, flexor surface of distal part of third pereiopod; d, lateral view of cephalic region; e, lateral view of second through fifth abdominal pleura; /, sternum of sixth abdominal segment and preanal carina; g, lateral view of preanal carina; A, lateral view of third pereiopod; i, flexor surface of distal part of fourth pereiopod; j, flexor surface of distal part of fifth pereiopod; k, dorsal view of telson. (Scales marked in 1 mm increments.)

spines. Sternum of fifth abdominal segment with lacking premarginal sclerotized spinules dorsally; prominent median tubercle, that of sixth about penultimate article approximately 1.4 times as 1.3 times as broad as long. Preanal carina with long as wide, its dorsal surface provided with well-developed spine. Telson about twice as long somewhat linearly arranged spinules. Coxae of as wide and with 7 to 9 spines in each of 2 dorsal third and fourth pereiopods lacking prominent rows. Antennular peduncle with proximal article ventrolateral spine. Third pereiopod with merus NUMBER 364 37 rounded ventrally, about 5.5 times as long as studded with paired, mesially concave rows of high, ventromesial surface bowed, never parallel corneous denticles (dextral row of 9 and sinistral to that of corresponding podomere of other third of 7) and posteromedian premarginal tubercle. pereiopod; lateral surface with small, inconspi- Basal segment of antennule (Figure \bb) with cuous, nonsclerotized tubercles; propodus about well-developed stylocerite overreaching mid- 5 times as long as broad; extensor and flexor length of segment, dorsal surface with clusters of surfaces bearing many small sclerotized spines; setae but lacking corneous spinules; distal margin dactyl approximately 3 times as long as wide, bearing row of 8 (left) and 6 (right) cornified freely movable, and bearing 2 oblique rows of spinules; penultimate segment of peduncle about sclerotized spines on flexor surface; rather prom- 1.4 times as long as broad, studded dorsally with inent band of setae extending from dorsolateral 7 (left) or 5 (right) small corneous spinules and 7 surface of basis to beyond midlength of lateral (left) or 8 (right) on distal margin; ultimate seg- margin of propodus. ment, about 0.5 as long as penultimate, armed HOLOTYPIC FEMALE.—Rostrum (Figure 15a,rf) with 7 spinules at base of lateral flagellum and with convex margins bearing trace of emargina- row of 4 at base of mesial flagellum. Antenna tion delimiting base of acumen, latter falling far with ventrolateral spine on basis almost reaching short of distal margin of proximal podomere of level of tip of stylocerite; lateral spine of scapho- antennule; rounded dorsal carina not excavate cerite weak, extending short distance beyond base dorsally (not dipping below level of lateral cari- of ultimate podomere of antennular peduncle and nae posterior to acumen) and ending preapically almost attaining distal end of that of antenna, on acumen; ventral carina lacking teeth or ser- lamella of scaphocerite overreaching peduncles of rations; ocellar beak, hidden in lateral view by both antennule and antenna. Flagellum of an- eyes, with anterodorsal margin arched, its ante- tenna broken, but that of paratype extending rior extremity falling short of faceted part of eye. beyond caudal margin of telson. Antennal spine acute; pterygostomian angle Third maxilliped overreaching antennular pe- acute but dorsal side not strongly concave, thus duncle by about 0.25 length of distal podomere; not appearing produced in prominent spine. Car- tip of exopod reaching base of apical third of apace devoid of other spines or ridges; surface penultimate segment of endopod; penultimate punctate with fine setae, lacking pile of short stiff segment of endopod about 1.4 times as long as setae. ultimate. Pleura of second and third abdominal segments First pereiopod slightly overreaching distal ex- (Figure 15*) with rounded posteroventral extrem- tremity of antennular peduncle; second over- ities, although fourth and fifth subangular, none reaching first by about 0.25 length of fingers; produced in spines. Ventral margin of third, terminal brush of setae lacking scraping denticles. fourth, and fifth pleura with linear clusters of 9, Third pereiopod (Figure 15cJi) unarmed except 6, and 4 fine corneous spinules, respectively, also for rather prominent spine on distal ventrolateral with additional fine plumose setae. Fourth ab- surface of merus, and when extended anteriorly dominal tergum approximately 0.9, 0.7, and 0.55 overreaching antennular peduncle by length of times as long as fifth, sixth, and telson, respec- propodus and dactyl. Merus with ventromesial tively; that of sixth about 1.3 times length of fifth margin weakly bowed, about 5 times as long as and 1.4 times that of telson. Sternum of fifth high, 2.3 times as long as carpus, and 2.1 times as abdominal segment (Figures \c, 15/) with rather long as propodus; latter about 4 times as long as large tuberculiform prominence, that of sixth wide and 1.1 times as long as carpus; distoventral about 0.76 as long as broad. Preanal carina (Fig- margin of coxa entire and both mesial caudo- ure \bg) with strong spine extending caudally ventral prominence and distolateral spine lack- beyond base of carina. Telson (Figure Ibfc) almost ing. Lateral, dorsal, and ventral surfaces of merus twice (1.9) as long as wide, its dorsal surface studded with weak, squamous tubercles flanked 38 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY distally by plumose setae; longitudinal band of and the fragmentary paratypic female, in the long plumose setae present on lateral surface; National Museum of Natural History, Smithson- mesial extremity of podomere produced in prom- ian Institution, number 184857. inent rounded lobe at level of mesial articular TYPE-LOCALITY.—Cole's Cave, Barbados, West condyle of carpus. Carpus studded with small, Indies, 13°10'40"N, 59°34'30"W. Notes accom- distally directed, corneous spines, and lateral sur- panying the holotype state that the specimen was face with prominent distoventral spine somewhat collected obscured by conspicuous ventrolateral tufts of high up the side of the large cavern. Water issues . .. through plumose setae extending along length of podo- an apperture [sic] estimated two yards horizontally wide mere, setae increasing in length from proximal to with 9" running sheet of water under 6" air space . . . distal end of podomere. Propodus also studded Shrimps seen 15' downstream from tunnel and made off to with corneous spines, those on flexor surface tend- the tunnel fast ... in the dark it took up a normal "legs down" position. In the light it went on its side and even on ing to be linearly arranged toward lateral surface its back and jerked about in spasms. which provided with conspicuous longitudinal band of plumose setae. Dactyl movable, its flexor The identity of the collector is unknown, but the surface bearing 9 corneous spinules arranged in specimen was obtained on 6 October 1972. The 2 somewhat arched rows; clusters of setae flanking paratypic female was found in the same cave by rows distally. R.R. Allen and C.E. Ray on 28 February 1963. Fourth pereiopod with dactyl reaching end of DISTRIBUTION AND SPECIMENS EXAMINED.— proximal third of carpus of third; merus almost Known only from the type-locality. The sources twice as long as carpus, latter about 0.7 as long as of the two known specimens are related under propodus. Fifth pereiopod reaching slightly be- "Type-locality": 19 (16.0), 23 Feb 1963, R.R. yond midlength of carpus of fourth pereiopod; Allen, C.E. Ray; 1$ (15.7, holotype), 6 Oct 1972. merus 1.3 times as long as carpus, latter approx- ECOLOGICAL NOTES.—See "Type-locality." imately 0.7 as long as propodus. Ornamentation LIFE HISTORY NOTES.—None. of merus, carpus, and propodus of fourth pereio- RELATIONSHIPS.—Atya brachyrhinus has its closest pod consisting of distal ventrolateral spine on affinities with the sympatric Atya innocous, with merus, ventrolateral row of 4 spines (increasing which it shares most of the features mentioned in in size distally) on carpus, and band of plumose the "Diagnosis" of each of the two. It differs from setae extending on lateral surface of podomeres all members of the genus in possessing a broad, from base of merus almost to end of propodus, short rostrum that has strongly convex lateral broadening distally on both merus and carpus. margins and does not reach the distal end of the Ornamentation of fifth pereiopod like that on proximal podomere of the antennule. It differs fourth except merus provided with 3 additional from A. innocous most conspicuously by the shape spines ventrally, and ventrolateral surface of car- of the rostrum; in that species the base of the pus with only 2 spines. acumen is clearly delimited by abrupt constric- Diaresis of lateral ramus of uropod flanked tions of the rostral margins. The pterygostomian proximally by row of 22 articulated corneous angle of A. brachyrhinus is less acute, the tubercles denticles and fixed spine at lateral end of row. on the third pereiopod are much less strongly COLOR NOTES.—On the label accompanying developed, and the spine on the preanal carina the holotype was the following: "It was colorless projects much beyond the basal part. From its with pink tail when seen. Two black dots only other sympatric relative, A. scabra, it differs more give away." strikingly by the absence of an angular bend in SIZE.—Carapace length of holotype and para- the rostral margin marking the base of the acu- typic female 15.7 and 15.8 mm, respectively. men, in the absence of corneous spinules on the TYPES.—The holotypic female is in the British dorsal surface of the proximal podomere of the Museum (Natural History), number 1972:539, antennule, in the absence of short stiff setae stud- NUMBER 364 39 ding the surface of the carapace, in the less spi- species were included. Doflein (1900), who fol- niform pterygostomian angle, in lacking corneous lowed Ortmann in the subgeneric assignment, denticles on the ventral margin of the pleuron of listed this shrimp from the Atlantic side of Pan- the second abdominal segment, in the proportion- ama. In the meantime, Ortmann (1897) had ately narrower telson with a greater number (7 or employed the combination Atya {Evatya) crassa, 8) tubercles in each of the dorsal rows, in the presented a diagnosis of the species, and included proportionately longer sternum of the sixth ab- Nicaragua and Mexico in his summary of distri- dominal segment, in the less well-developed tu- bution. Sheldon (1905) added no new informa- bercles on the lateral surface of the merus and tion. Bouvier (1905), not recognizing the subge- carpus of the third pereiopod, and in the size and neric designations of Ortmann, provided a key to arrangement of the tubercles on the flexor surface the species of the genus, and pointed out the of the propodus and dactyl of the third pereiopod. similarity in the serrate rostra of A. crassa and A. (= Micratya) poeyi Guerin-Meneville (1855). In his monograph of the Atyidae, Bouvier (1925) em- Atya crassa (Smith) phasized the relationships of A. crassa with A.

FIGURES lm, 9, 10, llg, 16, 17 gabonensis, considering them to be the most ad- vanced members of the genus. A key to the species Evatya crassa Smith, 1871:95-97 [type-locality: "Fresh water was provided along with a summary of the local- streams, Polvon, and the 'Rio Fulva, two and a half miles ity records (those just cited), an illustration, in northwest of Realejo,' Occidental Department, Nicara- gua"; types: (?)USNM 84261, 1?; (?)PM 6038, 16\ 1$].— lateral aspect, of a male, and another of the Kingsley, 1878b:57.—Holthuis, 1955b:26. telson. Pesta (1931), reporting on the Austrian Euatya crassa.—Koelbel, 1884:318,320. expedition of 1930 to Costa Rica, added a new Euatya (Evatya) crassa.—Koelbel, 1884:318. locality for the species: that flows into Atya (Evatya) crassa.—Ortmann, 1895:408, 410, 415.—Do- Grolfo Dulce. Oliveira (1945), in his study of Atya flein, 1900:127. scabra in northeastern Brazil, mentioned A. crassa Atya (Euatya) crassa.—Ortmann, 1897:184, 186. Evatya crassus.—Sheldon, 1905:343 [erroneous spelling]. but added nothing to our knowledge of the spe- Atya crassa.—Bouvier, 1905:110, 113, 124; 1925:26-28, 293, cies, nor did Holthuis (1951); however, the latter 319-323, figs. 54, 68.—Pesta, 1931:173, 178.—Oliveira, (1955b) cited this shrimp as the type of Smith's 1945:178.—Holthuis, 1951:9; 1955b:26, fig. 9.—Balss, Evatya and included the illustration of the entire 1955, fig. 1050.—Kaestner, 1970, figs. 13-18f.—Burukov- animal that appeared in Bouvier's monograph. sky, 1974, fig.85 . Balss' (1955) figure was also taken from Bouvier, REVIEW OF LITERATURE.—Smith (1871) de- Kaestner's (1970) from Balss, and Burukovsky's scribed this shrimp along with Atya rivalis and A. (1974) from Holthuis. tenella from freshwater streams at El Polvon in the PUBLISHED ILLUSTRATIONS.—The only illustra- western part of Nicaragua; he also cited a second tions of this shrimp are a dorsal view of the telson locality for A. crassa, "Rio Fulva, two and a half and a lateral one of an entire animal included in miles northwest of Realejo." Kingsley (1878b) Bouvier (1925). The latter figure was reprinted added no new information but listed the species by Holthuis (1955b) and Balss (1955), and indi- and cited it from the west coast of Nicaragua. rectly by Kaestner (1970) and Burukovsky (1974). Koelbel (1884), in describing Euatya sculptilis (= DIAGNOSIS.—Cephalic region of carapace Atya gabonensis Giebel, 1875) and emending the strongly sculptured and bearing many rows of spelling of Evatya, assigned Smith's species to the corneous-tipped spines and tubercles. Rostrum same genus and reported its occurrence in Rio with margins tapering from base and studded Presidio (perhaps near Mazatlan, Sinaloa), Mex- with median dorsal row of corneous-tipped spines. ico. Ortmann (1895) recognized two subgenera of Pterygostomian angle produced in prominent the genus Atya and placed A. crassa in the mono- spine. Ventral margin of abdominal pleura lack- typic subgenus Evatya; no new data relative to the ing sclerotized denticles; caudoventral angle of 40 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY fourth and fifth pleura usually produced in them bearing 2 large, acute, corneous-tipped spines. Fifth abdominal segment with median spines. Entire surface of carapace densely punc- tubercle on sternum moderately to strongly de- tate, and cephalic half very ornate dorsally and veloped; sternum of sixth abdominal segment less dorsolaterally, bearing paired ridges studded with than half as long as wide. Preanal carina in form rows of corneous-tipped spines; ventrolateral of subconical spine directed caudally to caudo- punctations bearing conspicuous setae. ventrally. Telson 1.2 to 1.4 times as long as broad Pleura of first and second abdominal segments and bearing paired arched dorsal rows of 4 to 6 (Figure 16J) with rounded posteroventral extrem- spines. Antennular peduncle with dorsal surface ities, corresponding parts of third obtuse, and of proximal article lacking sclerotized spinules those of fourth and fifth acute with corneous proximal to transverse distal row; penultimate apices; anteroventral angle of sixth also with article 1.5 to 2.2 times as long as wide and bearing acute, corneous, caudoventrally directed tip. All dorsolateral longitudinal row of spinules. Coxa of pleura lacking corneous denticles on ventral mar- third and fourth pereiopods with prominent ven- gin, but fourth and fifth with row of prominent trolateral spine, that of third also with strong plumose setae. Fourth abdominal tergum approx- mesial caudoventral prominence. Third pereio- imately 1.2 times as long as fifth, length of sixth pod with merus subplane ventrally, 1.5 to 2.3 subequal to that of fifth and almost 0.9 as long as times as long as high, ventromesial margin sub- telson. Sternum of fifth abdominal segment (Fig- angular and parallel to that of corresponding ure Im) with very prominent, corneous-tipped, podomere of other third pereiopod, and lateral acute median spine directed caudoventrally; ster- surface bearing irregular rows of spines, mar.y of num of sixth segment about 0.42 times as long as which with flattened corneous extremities; pro- broad. Preanal carina (Figure 16A) represented podus 1.3 to 1.6 times as long as broad, studded by strong, caudally directed, corneous-tipped with scalelike tubercles on extensor surface and spine. Telson (Figure 166) about 1.3 times as long with few similar ones of flexor surface, latter as wide, its dorsal surface bearing paired rows of largely obscured by conspicuous tufts of long 6 corneous denticles and with posteromedian tu- plumose setae; dactyl apparently inflexibly fused bercle slightly overhanging caudal margin. with propodus and bearing 1 or 2 small scalelike Proximal podomere of antennule (Figure 16a,/) tubercles on flexor surface just proximal to cor- with strong stylocerite overreaching midlength of neous tip. segment, dorsal surface with few setae but lacking MALE (Rio Chucunaque, Darien Province, corneous spinules; distal margin studded with Panama).—Rostrum (Figure I6a,c) with margins row of 5 corneous spinules; penultimate segment tapering from base to corneous, acute apex, latter of peduncle 1.5 times as long as wide and pro- almost reaching distal end of antennular pedun- vided with 4 spinules on dorsolateral surface and cle; dorsal median carina rounded and bearing row of 3 (right) or 4 (left) on distal margin; row of 7 corneous-tipped spines, 4 of which situ- ultimate podomere with no spinules on dorsal ated behind orbit, directed dorsally, and decreas- surface but with row of 5 flanking dorsal base of ing in size posteriorly; those on rostrum subequal lateral flagellum and 2 at dorsal base of mesial in size, 2 more posterior ones directed dorsally, flagellum. Antenna with ventrolateral spine on and anteriormost anterodorsally. Ventral surface basis reaching proximal end of penultimate pod- of rostrum evenly rounded transversely, lacking omere of antennular peduncle, distinctly over- clearly defined carina; lateral carinae very small, reaching stylocerite; lateral spine on scaphocerite becoming obsolete at base of anteriormost spine. strong, extending to level of tip of rostrum; la- Ocellar beak upturned and so well concealed mella overreaching peduncles of antennule and beneath rostrum and between eyes that hardly antenna; flagellum of antenna reaching caudal noticeable. Antennal and pterygostomian spines margin of telson. acute, and cephalic border of carapace between Third maxilliped overreaching antennular pe- NUMBER 364 41 (Ml

FIGURE 16.—Atya crassa (all from male from Rio Chucunaque, Panama, except b from male from Rio Lempa, El Salvador): a, dorsal view of cephalic region; b, dorsal view of telson; c, lateral view of cephalic region; d,ej, flexor surface of distal part of third, fourth, and Fifth pereiopods, respectively; g, ventral view of base of right third pereiopod; A, mesial view of appendices masculina and interna; t, lateral view of second through fifth abdominal pleura; j, lateral view of third pereiopod; k, lateral view of preanal carina; /, dorsal view of antennular peduncle. (Scales marked in 1 mm increments.) 42 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY duncle by 0.25 length of ultimate segment of and propodus of fourth pereiopod similar to that endopod; tip of exopod almost reaching mid- of third except ventral surface of merus with 3 length of penultimate podomere of endopod; pe- articulated spines, and ventrolateral surface of nultimate segment about 1.6 times as long as carpus with 1; only 2 such spines present on ultimate. merus of fifth pereiopod, and flexor surface of First pereiopod reaching level of base of ulti- dactyl with row of many more spinules, ornamen- mate podomere of antennule, second reaching tation otherwise similar to that of fourth. base of distal fifth of dactyl of first pleopod; Diaresis of lateral ramus of uropod flanked terminal brush of setae of both appendages lack- proximally by 16 articulated corneous denticles ing scraping denticles. Except for coxa, third and slightly larger fixed lateral spine. pereiopod (Figure I6dg,j) lacking spines and SIZE.—The largest specimen for which mea- when extended anteriorly, overreaching anten- surements are available is a male from Rio Pres- nular peduncle by length of propodus and dactyl. idio, Mexico, which has a carapace length of 60.0 Merus with ventromesial margin straight, almost mm. The largest female, one from northwestern 1.5 times as long as high, 14.1 times length of Ecuador, has a corresponding length of 41.4 mm. carpus, and 4.4 times as long as propodus; latter No ovigerous females have been reported or ex- slightly longer than wide and 0.48 as long as amined by us. carpus; distoventral margin of coxa strongly scal- DISTRIBUTION AND SPECIMENS EXAMINED.—This loped, distolateral surface with strong spine, and species ranges along the Pacific slope from Presi- mesial caudoventral prominence strongly devel- dio (probably Sinaloa), Mexico, southward to oped and studded with prominent setal clusters. Ecuador, and in Panama it has been found on Lateral, dorsal, and ventral surfaces of merus the Caribbean slope (Doflein, 1900) (Figure 17). bearing longitudinal rows of conspicuous, cor- Records for the known localities are listed be- neous tubercles of which apices of most somewhat low. Collections that we have examined are flattened and bearing sharp free edge; paired marked with an asterisk if they have been previ- clusters of plumose setae flanking distal base of ously reported and with a dagger if they are most tubercles; tubercles on ventral surface reported herein for the first time. Numbers follow- largely concealed by dense, shaggy beard of setae; ing the specimens listed are measurements, in mesial extremity of podomere slightly produced at level of mesial articular condyle of carpus; strong tubercle on distal mesioventral angle op- posing tubercle on carpus. Ventral and ventrome- sial surface of carpus heavily bearded. Flexor surface of propodus with few widely spaced tu- bercles, almost all obscured by tufts of setae borne proximally on podomere; dactyl fused with pro- podus, its flexor surface bearing median cluster of small, corneous denticles flanked by pair of setal clusters. Fourth pereiopod with dactyl reaching base of distal third of merus of third pereiopod; length of merus slightly greater than twice that of carpus, latter 1.1 times longer than propodus; coxa with strong distolateral spine. Fifth pereiopod reaching end of basal fifth of carpus of fourth; merus about 1.5 times as long as carpus, and latter 0.86 as long as propodus. Ornamentation of merus, carpus, FIGURE 17.—Distribution of Atya crassa. NUMBER 364 43 mm, of the carapace length or, if followed by Ecuador differ from those from the Middle Amer- "t.l.," total length. Some listings lack dates and/ ican region in possessing fewer tubercles on the or collectors; if so, these could not be determined. lateral surface and dorsal margin of the merus of MEXICO: (1) *BM, Rio Presidio (PSinaloa) the third pereiopod. In the South American spec- (Koelbel, 1884:318), 16* (60.0), 29 (17.7, 23.3), A. imens, the dorsal row consists of 12 to 14 (rarely Forrer; MHNP, 1 specimen (35.5). (2) fUSNM, 13 or 14) and in those from Middle America, 15 Rio Tehuantepee, Oaxaca, 16* (50.5), T. Mac- to 19. Six to 10 spines constitute the median row Dougall. on the rostrum, and, whereas in the specimens EL SALVADOR: fUSNM, Rio Lempa at Suchi- from Ecuador, Colombia, and Panama there are toto, 26* (29.3, 28.7), 1$ (29.0), 9 Feb 1924, C.A. usually six or seven, as contrasted with eight to Hildebrand and Foster; USNM, 26 (22.6, 32.8), 10 in the more northern part of the range, the 5 Feb 1924, CAH and F. numbers are not consistent, for specimens from NICARAGUA: (1) *USNM, freshwater streams, El both El Salvador and Nicaragua have as few as Polvon (12°27'N, 87°05'W) or Rio Fulva, 2.5 mi seven and at least one from each of Colombia NW of Realejo (Smith, 1871), 1$ (40.0), syntype, and Panama exhibit eight or nine. Of the other 1869, J.A. McNeil. (2) fPM, "Rio Frilo," (prob- features compared in specimens from throughout ably Rio Frio), about 1.8 km NW of Realejo, \6 the range, there seems to be no correlation of a (11.2), 2$ (8.5, 15.0), 1869, J.A. McNiel. variation with a limited part of the area occupied COSTA RICA: RIO Nuevo (Halbinsel Osa), Golfo by the species. Dulce (Pesta, 1931:178), 16* (120, t.l.), 4 juv, 1 ECOLOGICAL NOTES.—The only data available Apr 1930. as to the habitat occupied by Atya crassa are in PANAMA: (1) "Atlantic side" (Doflein, 1900: the short statement of Smith (1871:97), "Fresh 127). (2) fUSNM, Rio Chucunaque near Rio water streams ... and 'Rio Fulva' ... ," citing Sanson, Provincia de Darien, 36* (47.5,50.8,56.3), the localities from which his specimens were 27 Apr 1958, C.E. Bennett, Jr. (3) fUSNM, Rio taken, and the inclusion of "Rio" among the data Chucunaque above Membrillo, Provincia de Dar- accompanying most subsequent collections (see ien, 16* (38.9), 4$ (15.3-38.6), 7 Apr 1924, J.L. "Distribution"). Some of the specimens cited Baer. (4) fLGA, Rio Bayano at ford, Provincia herein from Ecuador were collected at an altitude de Panama, 19 (6.3), 6 Mar 1973, R.L. Dressier. of 150 m. (5) fLGA, headwaters of Rio Bayano 16* (28.0), LIFE HISTORY NOTES.—There are no published 19 (9.6), Dec 1974, L.G. Abele. data relative to the life history of members of this COLOMBIA: fUSNM, Rio Dagua at bridge, 0.25 species. Of the specimens known to us, not one is km from Buenaventura, 66* (30.8-41.0), 5$ ovigerous, and even the dates collections were (28.7-39.3), 17 Jul 1939, Karl P. Schmitt. made are few. Among the previously recorded ECUADOR: (l)fBM, NW Ecuador, alt 136 m, 26* localities, only that in Costa Rica (Pesta, 1931) (38.1, 39.0), 4$ (32.1-41.4). (2) fMCZ, Rios Cay- was accompanied by the date on which the col- apas, Hoja Blanca, and San Miguel, in immediate lection was made, 1 April 1930. vicinity of village of San Miguel, Provincia de REMARKS.—Fortunately, Atya crassa is the most Esmeraldas, 2$ (32.9, 37.9), Miyata and Rand. distinctive member of the genus, for one cannot (3) fUSNM, junction of Rio Cayapas and Rio be certain that any of the existing specimens San Miguel, Provincia de Esmeraldas, 19 (37.2), collected by McNiel in Nicaragua were among Jun 1977, Andris Rankis. the specimens on which Smith based his descrip- VARIATIONS.—The following few remarks are tion. The female in the Smithsonian (USNM based on such a limited series of specimens, 11 84261) that was formerly in the Kingsley collec- constituting the greatest number from a single tion is likely one of Smith's specimens. It bears a locality, that they should not be considered as hand written label carrying the following: "Eva- conclusive. The specimens from Colombia and tya Crussa [the "u" probably an intended "a"] 44 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Smith (type) for Locality s II & III Ann. Rep. spinules dorsally; penultimate article about twice Peabody Aca. Science," and number "216" is as long as wide, its dorsal surface usually provided included on a second small slip of paper. The with linear series of spinules. Coxae of third and smaller male and two females in the Peabody fourth pereiopods lacking prominent anterolat- Museum (no. 6083) from "Rio Frilo" are perhaps eral spines. Third pereiopod with merus rounded less likely to have been examined by Smith, for ventrally, more than 5 times as long as high; this stream is not mentioned at the end of the ventromesial surface bowed, never parallel to that description as is Rio Fulva. Inasmuch as the latter of corresponding podomere of other third pereio- specimens were collected within such a short dis- pod; lateral surface with linear series of subsqua- tance from Realejo, they are at least virtual to- mous, comparatively large tubercles, those con- potypes. stituting median series rather larger than others but none heavily sclerotized; propodus 3 to 4 Atya dressleri Abele times as long as broad, extensor surface bearing rather large subsquamous tubercles, flexorsurfac e FIGURES Ig, 9, 10, 12/, 18, 19 also with large tubercles bearing sclerotized Atya dressleri Abele, 1975:51-57, figs. 1, 2 [type-locality: Rio spines, more lateral tubercles forming row and all San Juan, 15 km above town of Calobre (566 m elevation), flanked by horseshoe-shaped row of setae (latter Provincia de Veraguas, Panama; types, USNM 184856, sometimes absent in later intermolt stage); dactyl 6* holotype; USNM 141845, 8$ paratypes, 16* paratype]. about twice as long as broad, freely movable, and REVIEW OF LITERATURE.—Abele's original de- usually bearing 2 oblique rows of sclerotized scription and notes on the range and habitat (see spines on flexor surface, occasional members below) of this Panamanian species constitute the slightly displaced from either series and rarely only published record of its existence. spines arranged in single row; except in middle PUBLISHED ILLUSTRATIONS.—The illustrations to late intermolt individuals, conspicuous band of accompanying the original description include setae extending from basal third of lateral surface dorsal and lateral views of the cephalic region, of merus to distal margin of propodus. dorsal view of the telson, illustrations of the an- HoLOTYPic MALE.—Rostrum (Figure I8a,d) tennule, third pereiopod, first pleopod of the with margins tapering from base, and hardly hint male, appendices interna and masculina, preanal of angle at base of long acumen; apex of latter carina, and a lateral view of an entire animal slightly overreaching extremity of proximal pod- showing the color pattern. omere of antennular peduncle; dorsal median DIAGNOSIS.—Cephalic region of carapace not carina gently curved, not excavate dorsally (not conspicuously sculptured, glabrous, lacking dipping below level of lateral carinae posterior to spines other than antennal, pterygostomian, and acumen), and reaching apex of acumen; ventral usually ventral rostral; pterygostomian spine ab- carina with 2 preapical teeth, both on acumen; sent or rather small; rostrum with margins taper- ocellar beak well hidden between eyes, reaching ing from base, sometimes with slight irregularity little beyond level of base of stylocerite, its ce- but never angulate. Ventral margin of all abdom- phalic border broadly rounded, and dorsal mar- inal pleura without sclerotized denticles, and cau- gin embraced by sides of ventral rostral groove. doventral angles of fourth and fifth pleura not Antennal spine strong; pterygostomian spine produced in spines. Sternum of fifth abdominal rather small; no spine present between them. segment with small median tubercle, that of sixth Surface of carapace bearing crowded minute 1.4 to 1.5 times as broad as long. Preanal carina punctations supporting very short, erect, fine se- with well-developed curved spine. Telson almost tae, latter nowhere conspicuous; devoid of ridges twice as long as wide and with 8 to 10 spines in and spines other than those just mentioned. each of 2 dorsal rows. Antennular peduncle with Pleura of first 3 abdominal segments (Figure proximal article lacking premarginal sclerotized 18*) with rounded posteroventral extremities; cor- NUMBER 364 45

FIGURE 18.—Atya dressleri (all from holotypc except k,l from Santa Fe, Panama): a, dorsal view of cephalic region; b, dorsal view of antennular peduncle; c, mesial view of appendices masculina and interna; d, lateral view of cephalic region; e, lateral view of second through fifth abdominal pleura; /, flexor surface of distal part of fourth pereiopod; g, subflexor surface of distal part of fifth leg; A, dorsal view of telson; i, lateral view of preanal carina; j, lateral view of third pereiopod; k,l,m, flexor surface of distal part of third pereiopod. (Scales marked in 1 mm increments.) 46 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY responding parts of fourth and fifth angular but reiopod (Figure 18/,/n) without lateral distoven- not produced in spines. All pleura lacking cor- tral spine on merus and carpus, ventral spine neous spinules on ventral margin, but fourth and absent from merus, and carpus lacking distola- fifth with moderately prominent fringe of plu- teral spines; when appendage extended ante- mose 'setae. Fourth abdominal tergum slightly riorly, overreaching antennular peduncle by dac- more than 1.1 times as long as fifth, 1.2 times as tyl, propodus, and distal third of carpus. Merus long as sixth, and tergum of fifth subequal in with ventromesial margin bowed, almost 7 times length to telson. Sternum of fifth abdominal seg- as long as high, 2.6 times as long as carpus, and ment with small acute median tubercle (Figure 2.3 times as long as propodus; propodus 3.4 times Ig). Sternum of sixth abdominal segment about as long as wide and almost 0.9 as long as carpus; 0.7 as long as broad. Free part of preanal carina distoventral margin of coxa entire (evenly (Figure 18i) spiniform, curved, and slightly ov- rounded), and mesial caudoventral prominence erreaching caudal border of basal part of sclerite. absent. Lateral, dorsal, and ventral surfaces of Telson (Figure 18A) .little less than 1.6 times as merus studded with moderate number of rather long as broad, its dorsal surface bearing paired, large tubercles little if any cornified and, for most concave rows of 7 corneous denticles flankedpos - part, arranged in linear series; longitudinal band teriorly by 1 (right) or 2 (left) ones; posterome- of setae present on lateral surface of merus; mesial dian tubercle slightly overhanging caudal mar- extremity of podomere noticeably produced in gin. rounded lobe at level of mesial articular condyle Proximal podomere of antennule (Figure 18b) of carpus. Latter strongly tuberculate, almost all with stylocerite reaching between base of distal tubercles flanked distally by subsemicircular row fourth and fifth of segment; dorsal surface with of setae; lateral and ventrolateral surfaces matted linear cluster of setae but lacking corneous spi- with tufts of plumose setae. Propodus also nules; distal margin bearing row of 9 (right) or 6 strongly tuberculate, almost all tubercles with (left) corneous spinules; penultimate segment of small cornified spines and, like those of carpus, peduncle slightly less than twice as long as wide flanked by setae; those on flexor surface not in and bearing dorsal longitudinal row of 7 corneous well-defined rows proximally; lateral surface with spinules and transverse distal row of 6; ultimate tufts of setae arranged in longitudinal band. Dac- podomere without spinules on dorsal surface and tyl movable, its flexorsurfac e with 2 oblique rows with row of 7 (right) or 8 (left) at base of lateral of denticles flanked distally by usual setal clusters. flagellum and another of 2 at mesial base of Fourth pereiopod with dactyl reaching end of mesial flagellum. Antenna with ventrolateral proximal fourth of propodus of third pereiopod; spine on basis reaching almost as far anteriorly as length of merus slightly more than twice as long stylocerite; lateral spine on scaphocerite rather as carpus; latter approximately 0.9 as long as strong, reaching about midlength of ultimate propodus. Fifth pereiopod reaching anteriorly to podomere of antennular peduncle; lamella far end of proximal third of propodus of third pe- surpassing latter; flagellumextendin g to fifth ab- reiopod; merus 1.4 times as long as carpus, latter dominal tergum. almost 0.7 as long as propodus. Ornamentation Third maxilliped overreaching antennular pe- of merus, carpus, and propodus of fourth pereio- duncle by distal fifth of distal podomere of en- pod consisting of distolateral spine and 2 more dopod; tip of exopod attaining base of distal fifth proximal ones on ventral surface of merus (more of penultimate podomere. distal spine of latter very long), distal ventrolat- First pereiopod reaching level of base of ulti- eral spine and 2 small distolateral spines on car- mate podomere of antennular peduncle, second pus; lateral surface of merus, carpus, and propo- extending to base of distal fifth of fingers of first dus with conspicuous band of plumose setae. pereiopod; terminal brush of both appendages Ornamentation of corresponding podomeres of lacking setae with scraping denticles. Third pe- fifth pereiopod quite similar to that of fourth NUMBER 364 47 except distalmost of ventral spines on merus not Panama and Veraguas provinces at elevations of nearly so large. 566 to 650 m (Abele, 1975:56) (Figure 19). Diaresis of lateral ramus of uropod flanked Records for the known localities are listed be- proximally by row of 21 articulated corneous low. Collections that we have examined are denticles, and fixed spine at lateral end of row. marked with an asterisk if they have been previ- COLOR NOTES (paraphrased from Abele, ously reported and with a dagger if they are 1975:55, 56).—Ground color light brown with reported herein for the first time. Numbers follow- yellow and black specks. Carapace light brown ing the specimens listed are measurements, in with dark brown oblong area on anterolateral mm, of the carapace length. Some listings lack surface just dorsal to elongate black spot. Short the date the collection was made; this could not yellow rectangular area with black posterior bor- be determined. der posterodorsal to latter. Posteroventral part of PANAMA: Provincia de Panama—(1) Rio Cabra carapace with long yellow stripe margined in above Cerro Azul, alt 650 m (Abele, 1975), 26 black. First abdominal segment with sinuous ("molts of aquarium specimens"). (2) fLGA, black area covering anterior part of segment; Cerro Jefe, Pacora Basin, Id* (exuviae only), 1969, black area bordered by distinct narrow yellow R.L. Dressier. (3) fUSNM, stream between Pa- band. Second abdominal segment with anterior cora and Chepo, 17 km E of Rio Pacora, 16* and posterior pleural angles outlined in black and (12.2), 1 exuviae, 20 Feb 1971, RLD. Provincia de with oblique black stripe medial to each; distinct Veraguas—(A) *USNM, "Probably near 'Goofy yellow area with black border situated postero- Lake,'" 1<5 (exuviae, 15.1), RLD. (5) *USNM, dorsally. Third segment with pleural angle black trib of Rio Santa Maria N of Santa Fe, alt 600 m and bearing oblique stripe anterior to it. Fourth (Abele, 1975),6$paratypes (10.7, 14.1, 16.4, 17.0, segment bearing hourglass-shaped yellow area 17.0, 18.2), 9 Feb 1962, H. Loftin and E.W. bordered in black on posterodorsal surface; Tyson. (6) *USNM, headwaters of Rio San Juan pleural angle black and oblique black stripe lying about 15 km above Calobre, alt 566 m (Abele, anterior to it. Pleural angle of fifth segment like 1975), 2d (20.0, paratype; 20.4, holotype), 2$ that of fourth. Almost entire dorsal surface of sixth segment yellow with black border. Distinct black spot present at base of uropods. Antennular peduncle with dorsal surface of antepenultimate and penultimate segments bearing yellow mark- ings; flagella brown. Third through fifth pereio- pods with 7 yellow bands: 1 at coxa-ischium, 1 about midlength and another toward distal end of merus, 1 at merus-carpal joint, 1 distally on carpus, and 1 at each end of propodus. Remain- der of shrimp concolorous brown. Some individuals lack, or have a modification of, the pattern described; others have a mid-dor- sal longitudinal stripe; and the smaller specimens were concolorous light brown. SIZE.—The carapace length of the males ranges from 12.1 to 20.4 mm, and that of the females, from 9.3 to 18.4 mm. No ovigerous females are available. DISTRIBUTION AND SPECIMENS EXAMINED.— Known only from the Pacific slope of Panama in FIGURE 19.—Distribution of Atya dressleri. 48 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY paratypes (9.3, 18.4), 20 Feb 1973, L.G. Abele Atya sculptata Ortmann, 1890:465, 466 [type-locality: Africa, and H. Robinson. "Vielleicht aus West-Africa ... "; type^3trassburg Mu- seum].—Aurivillius, 1898:14.—OH veira, 1945:179. VARIATIONS.—The most conspicuous variations [Atya] sculptipes.—Ortmann, 1890:466 |erroneous spelling]. occur in the spination of the rostrum, the ventral Atya (Atya) gabonensis.—Ortmann, 1895:410, 414, 415 [by margin bearing one to three spines or serrations, implication]; 1897:184, 185, 186 [by implication]. and, whereas the dorsal surface is usually entire, Evatya sculptilis.—Ortmann, 1897:185 [erroneous transcrip- the rostra of the exuviae of one specimen and of tion of Euatya]. a paratypic female exhibit a row of four serra- Atya scabra.—Rathbun, 1900:313, 314 [in part].—Balss, tions. The pterygostomian angle may be strongly 1914:98 [in part]. Atya.—Yaro, 1967:210.—Monod, 1967:176. reduced and rounded, or it may be moderately produced, ending in an acute spine. The telson REVIEW OF LITERATURE.—This shrimp was de- bears seven to 10 spines in each of the paired scribed in considerable detail by Giebel (1875) on dorsal rows. Although in most of the specimens the basis of four specimens that had been sent to the spines (usually about eight) on the flexor him by Baron von Koppenfels from Gabon. Al- surface of the dactyl of the third pereiopod are most a decade later, Koelbel (1884) presented a arranged in two rows, in two individuals exam- description and beautiful illustrations of an atyid ined they occur in a single median series or in a from the Orinoco Basin of South America that row with one or two spines flanking it. he designated Euatya sculptilis, employing a ge- ECOLOGICAL NOTES.—Abele (1975:56) stated neric name that is an invalid emendation of that this shrimp "occurs in small fast running Evatya Smith (1871:95). freshwater streams from about 560 to 650 m In recording the decapod crustaceans housed elevation. All of the specimens were collected in the Strassburg Museums, Ortmann (1890) in- from vegetation or overhanging roots in areas cluded a brief description of an African shrimp, where the current was swift." proposing the name Atya sculptata and assigning LIFE HISTORY NOTES.—The only information it to "die margaritacea-Gruppe (zu der margaritacea, concerning the life history of members of this robusta, scabra und sculptipes [sic] gehoren)" (p. species is that offered by Abele (1975:56): "This 466). In a subsequent study of the geographical species is fairly long lived. Robert Dressier has distribution of the family Atyidae, Ortmann kept adult individuals alive in an aquarium for (1895) placed Koelbel's Euatya sculptilis and his more than 5 years." Atya sculptata in the synonomy of Atya gabonensis and employed the subgenus Evatya in referring to Atya gabonensis Giebel Atya crassa (p. 184), thereby implying that the other members of the genus belong to the nomi- FIGURES In, 3, 9, 10, 11/, 20, 21 nate subgenus. In addition, he reviewed the Atya gabonensis Giebel, 1875:52-55 [type-locality: Gabon; ranges of A. gabonensis, A. crassa, and A. scabra. In types: disposition unknown].—Thompson, 1901:22.— his account of the freshwater shrimps of South Bouvier, 1904:138; 1905:110, 112, 123, 124, 128, fig. 26; America (1897), he included a diagnosis of this 1925:20, 22, 26, 27, 29, 293,317-323,357, figs.707 , 708.— Monod, 1928:205; 1933:461; 1967:110,119,135, pi. 7: fig. species and stated that it is possible that it occurs 8; pi. 9: figs. 25, 26; 1977:1203, 1204; 1980: 375— Oliv- only in the Orinoco and in Gabon. A single male eira, 1945:179.—Irvine, 1947:306, fig. 211.—Holthuis, specimen in the National Museum of Natural 1951:9, 25; 1980:69, 71, 181.—Gordon, 1967:52.—Reed, History collected by O.F. Cook at Mt. Coffee, 1967:120, fig. 153.—Motwani and Kanwai, 1970:34.— Liberia, provides evidence that Rathbun (1900) Rutherford, 1971:87, 90.—Lemasson, 1973:68, 70.—Pow- ell, 1979:116, 134-138.—Hobbs, 1980:111. perhaps had a specimen of A. gabonensis before her Euatya sculptilis Koelbel, 1884:317-321, 323, pi. 2: fig. 8; pi. in preparing her account of the decapod crusta- 3: figs. 1-8 [type-locality: Orinoco; type: NMW, 1 speci- ceans of West Africa; however, there was no men].—Hobbs, 1980:111. identification accompanying the uncatalogued NUMBER 364 49 specimen. She did not misidentify one of the tion of the shrimp and stated that this species is specimens of Atya scabra she reported from St. thought to be confined to rocky sections of the Paul's River near Mt. Coffee, Liberia (see "Dis- river (Volta). tribution" of A. scabra herein). She erred, however, In his review of the caridean Crustacea of in considering specimens of A. qfricana to be con- tropical West Africa, Holthuis (1951) provided a specific with A. scabra in synonymizing A. gabonen- nearly complete synonomy for the species and sis with the latter, listing the Orinoco and "Ga- listed all of the known locality records, expressing bun" as localities for A. scabra. Thompson (1901) doubt, however, of the validity of the reported reported the presence of Atya gabonensis among the occurrence of the species in the Orinoco River, Crustacea "contained in the Museum of Univer- and erring in treating Rathbun's (1900), and sity College, Dundee," but noted the absence of perhaps Johnston's (1906), records for A. scabra in locality data. Liberia as those of A. gabonensis. Monod (1967) In recording the specimens of atyids in the cited the range of this shrimp as extending from Museum d'Histoire Naturelle (Paris), Bouvier to Gabon; his illustrations were taken (1904) stated that this shrimp is the most beau- from Bouvier (1925) and Irvine (1947). Gordon tiful and largest species belonging to the genus (1967) included no information concerning this Atya. He listed specimens from three localities (see shrimp. Reed (1967) reported its presence in most "Distribution and Specimens Examined"). The parts of northern Nigeria but noted that it is not following year (1905) he elaborated upon infor- common; it inhabits rocky areas of the "main mation offered in 1904, grouped A. gabonensis with river" and its tributaries. Yaro (1967) recorded the advanced members of the genus, and pre- the common names in several of the local dialects. sented illustrations of the rostrum and cephalo- Motwani and Kanwai (1970) reported the oc- lateral region of the carapace. In his monograph currence of large numbers of Atya gabonensis in the of the family Atyidae, Bouvier (1925) first ques- coffer-dammed right channel of the Niger River tioned the reported source of Koelbel's specimens, on 4 and 5 August 1966. Gill nets ensnared them stating that "POrenoque, ce qui tient peut etre a at depths as great as 11.4 m. Rutherford (1971) une fausse indication de provenance" (p. 319). added no new information other than his failure Following a detailed description and a presenta- to find members of the species along the Cape tion of the same figures published in 1904, he Coast of Ghana. Lemasson (1973), noting that summarized the range, restated the sizes (45 to this shrimp attains a total length of 12.4 cm, 124 mm) of the specimens available to him, and believed that it might lend itself to commercial discussed the relationships, stating that the most cultivation. Monod (1977) reported the range of highly evolved stock of the genus "a produit en A. gabonensis to include the coastal streams of West Afrique occidentale et en Amerique VA. gabonensis Africa and the islands of Fernando Poo and Sao et dans la region de l'isthme americain VA. crassa Tome; also he suggested that the distribution of qui represente avec l'espece precedente le point the species may be associated with rocky sub- terminal" (p. 357). strates. Monod (1928 and 1967) listed this shrimp as Powell (1979) noted that A. gabonensis occurred being fished in Cameroon and later (1933) re- in company with Potamalpheops haugi (Coutiere, ported it from the "Yabassi riv. Wouri," Came- 1906) in "whitewater mollusc rivers" (p. 116) of roon. In updating our knowledge of the crusta- the Niger Basin in Nigeria. Juveniles were found cean fauna of West Africa, Monod (1933) pre- "in the Nun branch of the Niger near Kaiama sented a new locality record for A. gabonensis and (about 95 km inland and 15 km upstream from questioned its occurrence in the Orinoco. Oliveira the extreme dry-season tidal-limit) ..." (p. 134). (1945) added no new data. Another locality was He pointed out that the locality Bouvier cited by Irvine (1947) who provided an illustra- (1925:319) cited on the Ogooue River lies about 50 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

125 air-km from the coast. The similarity in body subconical and directed almost ventrally. Telson form of migrating juveniles of A. gabonensis and of 1.2 to 1.4 times as long as broad and bearing individuals of P. haugi of equal size render them paired arched dorsal rows of 4 or 5 spines. Anten- "difficult to separate with the naked eye" (p. nular peduncle with dorsal surface of proximal 136). Like P. haugi, according to Powell (p. 137), article bearing 1 to 6 spinules proximal to distal A. gabonensis "is perhaps restricted to relatively row; penultimate article 1.4 to 1.5 times as long hard waters," and he implied that it is present in as wide and bearing scattered sclerotized spinules the Osse watershed. Holthuis (1980) summarized in lateral half of dorsal surface. Coxae of third our knowledge of this shrimp, providing FAO and fourth pereiopods lacking prominent spine (Food and Agriculture Organization of the anterolaterally. Third pereiopod with merus United Nations) names, local names, geographi- somewhat flattened ventrally, about 2.5 times as cal and habitat distribution, and maximum size long as high; ventromesial surface parallel to that (total length: 124 mm for males and 92 mm for of corresponding podomere of other third pereio- females). References to its importance in fisheries pod, and lateral surface bearing irregular rows of are also included. Hobbs (1980:111) reported the heavy corneous tubercles with flattened, often presence of this species in Rio Piaui, Brazil. frayed, scalelike extremities; propodus 1.6 to 1.9 PUBLISHED ILLUSTRATIONS.—The first and best times as long as broad; extensor surface studded illustrations of Atya gabonensis are those of Koelbel with stalked corneous tubercles similar to those (1884) and include a beautifully executed lateral on merus and with very few widely spaced ones view of the entire body and appendages along on flexor surface, most tubercles on latter surface with a ventral view of the third pereiopod, lateral flanked distally by clusters of plumose setae; dac- view of the first pereiopod, and a posterior view tyl at least slightly movable, its flexor surface of each of the gnathal appendages. Bouvier (1905) bearing 1 to several tubercles in cluster just prox- provided a dorsal view of the rostrum, and the imal to corneous tip. rostral, orbital, and pterygostomian regions of the MALE (Volta River at Kpong Rapids, carapace were presented in lateral aspect. The Ghana).—Rostrum (Figure 20a,d) with margins latter figures were used by Bouvier (1925) in his somewhat tapering, but thickened and appearing monograph. The most recent illustration is that concave laterally, to base of acumen where pro- of Irvine (1947), a lateral view of the entire animal duced in spines reaching slightly beyond mid- that is much less detailed than Koelbel's figure 1 length of proximal podomere of antennule; apex on plate 3. Bouvier's figure has been reproduced of acumen overlying midlength of penultimate by Monod (1967) and Reed (1967). podomere of antennule; dorsal median carina, DIAGNOSIS.—Cephalic region of carapace although prominent, dipping below level of lat- strongly sculptured, but spines limited to anten- eral carinae posterior to acumen and reaching nal, pterygostomian, and occasional ventral and apex of latter; ventral carina poorly developed, often lateral rostral; pterygostomian spine rather ventral surface of rostrum grooved posteriorly, weak. Rostrum with margins suddenly con- receiving ocellar beak which strongly produced tracted, forming distinct angle anterior to orbit, anteriorly, ending in acute apex almost reaching angle often produced anteriorly; dorsal surface level of tip of stylocerite. Antennal and pterygos- without row of strong spines but with low median tomian spines moderately strong and acute, no carina. Ventral margin of abdominal pleura lack- spines present between them. Surface of carapace ing sclerotized spinules. Sternum of fifth abdom- densely punctate and rather ornate, marked by inal segment with median tubercle produced in strong ridges; submarginal punctations along curved hornlike projection usually overreaching ventral edge of carapace bearing conspicuous midlength of sixth sternite when abdomen flexed. setae. Sternum of sixth abdominal segment approxi- Pleura of first 3 abdominal segments (Figure mately one-half as long as wide. Preanal carina 20i) with rounded posteroventral extremities, cor- NUMBER 364 51

FIGURE 20.—Atya gabontnsis (all from male from Kpong rapids of Volta River, Ghana, except A from Rio Piaui, Brazil): a, dorsal view of cephalic region; b, mesial view of appendices masculina and interna; c, lateral view of preanal carina; d, lateral view of cephalic region; t, flexor surface of distal part of fourth pereiopod; /, lateral view of distal part of third pereiopod; gji, flexor surface of distal part of third pereiopod; i, lateral view of second through fifth abdominal pleura; j, dorsal view of telson; k, dorsal view of antennular peduncle; /, flexor surface of distal part of fifth pereiopod; m, lateral view of third pereiopod; n, ventral view of basal podomeres of right third pereiopod. (Scales marked in 1 mm increments.) 52 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY responding parts of fourth and fifth segments antennular peduncle, second reaching base of rounded to subangular but not produced. All distal fourth of fingers of firstpereiopod ; terminal pleura lacking corneous spinules on ventral mar- brush of setae of both appendages lacking scrap- gin but fifth with conspicuous row of plumose ing denticles. Third pereiopod (Figure 20fg,m,n) setae. Fourth abdominal tergum about 1.2 and lacking spines, and when extended anteriorly ov- 1.1 times as long as fifth and sixth, respectively, erreaching antennular peduncle by length of car- and subequal in length to telson; length of sixth pus, propodus, and dactyl; merus with ventrome- only slightly greater than that of fifth and 0.9 as sial margin straight, about 2.5 times as long as long as telson. Sternum of fifth abdominal seg- high and almost twice length of carpus; latter 2.3 ment with very large median, curved, hornlike times as long as propodus; length and width of projection (Figure In) which, when abdomen propodus subequal and its length 1.5 times that flexed, abutting sternum of sixth abdominal seg- of dactyl; distoventral margin of coxa very weakly ment immediately anterior to preanal carina. scalloped and mesial caudoventral prominence Latter sternum approximately 0.4 as long as studded with conspicuous setal clusters. Lateral, broad. Preanal carina (Figure 20

FIGURE 21.—Distribution of Atya gabonensis (circled numerals = number of localities). mately 6°29'N, 6°55'E (CBP, pers. comm.). (8) near Brokopondo (Holthuis, pers. comm.), 1$ Cross River at Itigi, 5°53'N, 8°01.5'E (GBP, pers. (27.5), 14 Feb 1964, M. Boeseman. comm.). According to C.B. Powell, it has not BRAZIL: *USNM, Rio Piaui, trib of Rio Par- been found in the Benin, Sombreiro, or New naiba, at Boa Esperanza Dam, state of Piaui Calabar rivers, nor are there records of its occur- (Hobbs, 1980:111), 56* (50.0-52.7), 31 May 1968, rence in Sierra Leone or Ivory Coast. J.W. Bezerra e Silva. CAMERON: Wouri River, Yabassi (Monod, NO DATA: *BM, label reads "?Type of Atya 1933:461). scabra Leach. Trans. Linns. Soc. Lond. xi p. 345," GABON: *BM, no locality, 16* (42.1). (1) 16* (38.8). *MHNP, Ngomo dans la riviere Ogooue (Bou- The presence of the species in Liberia (Holt- vier, 1925:319), 26* (12.6, 18.7), 3$ (19.0, 39.4, huis, 1951:25) is confirmed by the recent discov- 40.0), 1906, E. Haug. ery of an uncatalogued specimen from Mt. Coffee ZAIRE: (1) fRNHL, Rapids of lower Congo in the collection of the Smithsonian Institution. River at Kinsuka near Kinshasa, 1$ (43.4), 21 We have examined the specimens from the St. Jun 1971, P. Brichard. (2) fMCZ, Zaire River Paul River and from Beulah reported by Rath- near Inga Hydroelectric Dam, 5°31.5'S, bun (1900:313) as Atya scabra, and later cited by 13°37.5'E, 16* (45.3), 4 Aug 1973, T.R. Roberts, Holthuis as members of A. gabonensis, and we DJ. Stewart. concur in Rathbun's determinations. Her speci- FERNANDO POO: No locality (Monod, 1977: men from "Muhlenburg Mission," as pointed out 1204). above, is a member of Atya africana. The Johnston SAO TOME: NO locality (Monod, 1977:1204). (1906:862) record for A. scabra in Liberia (re- WEST AFRICA (probably Ghana): *BM, 36* ported as A. gabonensis by Holthuis, 1951:25) in- (31.5-42.0), FRI. cluded no locality and conceivably was based on MEXICO: fBM, 1 dry specimen (43.9). (We ques- Rathbun's (1900) report. In any event, to our tion data.) knowledge, there are no extant specimens of A. VENEZUELA: Orinoco River (Koelbel, 1884: gabonensis from Liberia except the male from Mt. 321). Coffee (USNM). SURINAM: *RNHL, rapids in Suriname River The record for Mexico cited here is also in need NUMBER 364 55 of confirmation. In the absence of more precise parable to the speed of darting fish" (p. 136). The locality data, we are not including Mexico within specimens collected from the Kpong Rapids of our summary statement of the range of the spe- the Volta River by Berner (see "Distribution and cies. Specimens Examined") were found "under rocks As for the specimen in the British Museum that three to five feet deep in swiftly flowing, murkey bears the questionable label "Type of Atya scabra," water." According to C. B. Powell (pers. comm.), we are convinced as was Holthuis (1966:234) that this shrimp is not confined to rapids in Nigeria. this specimen was not that (or one of those) before "The localities I know them from are wide deep Leach when he described Atya scabra. Supporting rivers with a large seasonal change in water level this conclusion is the illustration in plate 21 of and discharge—in the dry season the rivers are Leach's 1816 account of this species. sluggish, and in the Nun and lower Osse rivers VARIATIONS.—The variations noted by us are there are no rocks at all." hardly noteworthy, and none seems to be re- LIFE HISTORY NOTES.—The only ovigerous fe- stricted to populations occupying a restricted part males of which we are aware are the three (c.l. of the range of the species. The rostral margins 29.7, 30.0, and 30.6 mm) noted here that were may be almost straight or distinctly concave. collected from the Volta River in Ghana, on 20 Whereas the ornamentation of the cephalic and July 1950. The diameters of the eggs borne by ambulatory appendages exhibits variations in them range from 0.4 X 0.5 to 0.4 X 0.7 mm. The number and position of the spines and tubercles juveniles as noted in "Ecological Notes" have and in the condition of the setation borne by been found in loose vegetation, and they have not them, the differences are indeed minor. been reported as taken on rocky substrates as ECOLOGICAL NOTES.—The first note on the type have the adults. of habitat occupied by this shrimp was Bouvier's We were informed by C.B. Powell (pers. (1904:138) record "Chutes de Felou," and re- comm.) that during the low-water season (De- peated by him (1905:124; 1925:319). He added cember to April) in Nigeria, ovigerous females are (1925) that the specimens reported from "Ngomo, quite common. During the rest of the year, how- dans Ogooue" were captured from crevices in ever, collecting is not possible. He attempted to rocks on the river bed. That at least the adults of rear larvae obtained from females collected in the the species may well be largely restricted to rapids Osse River, but in fresh water they failed to molt is suggested by Irvine (1947:306) who stated that and died after about 10 to 12 days. With the it "is said to inhabit only those parts of the river addition of salt water, some of the larvae molted where the bottom is rocky." Reed et al. (1967:120) to the second instar. He suggested that the 10- also noted that the usual habitat is in rocky parts day survival period might allow the larvae time of streams, and Motwani and Kanwai (1970) to reach estuaries from such far inland localities reported the occurrence of this shrimp in an as Kainji. Juveniles that were placed in round impounded channel of the Niger River where it plastic containers swam continuously round and was taken in gill nets set at depths of as great as round the edge at remarkable speeds, and when 11.4 m. The most recent observations are those of blocked, they scrambled over the obstacle, leaving Powell (1979) who remarked that A. gabonensis the water, if necessary, to continue their course. occurs in "whitewater mollusc rivers" (p. 116), Powell suggested further that young Atya, like at "is perhaps restricted to relatively hard waters" least some young Macrobrachium, may avoid the (p. 137), and that juveniles had been found stronger currents by moving upstream close to "among loose vegetation and other debris caught the shore. in the current by parts of fallen trees projecting COMMON NAMES.—The following common out of the water" (p. 134). He also stated that names have been reported for this shrimp (no "migrating juveniles of Atya gabonensis . .. can distinction is apparently made between it and the swim steadily forward with surprising speed, com- African Atya scabra). 56 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Bomingomo (in Batanga), Cameroon (Monod, 1928:206) and Hobbs, 1969:61.—Abele, 1975:56, 57.—Abele and Crevettc gros-doigt (in French), Cameroon (Monod, Blum, 1977:240, 242-245, 250. 1928:458) Atya scabra.—Ortmann, 1895:409 [in part]; 1897:184 Dikuta (in Bassa Bania), Cameroon (Monod, 1928:206) [in part].—Rathbun, 1901:119 [in part].—Bouvier, Ekusa (in Soubou), Cameroon (Monod, 1928:206) 1905:121 [in part]; 1925:314 [in part]. Jaten lendi (in Hausa), in Nigeria (Yaro, 1967:210) Atya innocous.—Holthuis, 1966:237, 238; 1969:92; 1980:69, Kalama-toruopuru (in the Kolokuma dialect of Ijaw), 181.—Beatty, 1968:263.—Chace and Hobbs, 1969:5, 15, around Kaiama, Nigeria (C.B. Powell, pers. comm.) 19, 33, 36, 44, 46, 47, 57-62, 66, figs. 8, lOa-c, 14a,b, pi. Mobengomo (in Douala), Cameroon (Monod, 1928:205) 1.—Chace, 1972:14.—Bonnelly de Calventi et al., Ndakansa (in Nupe), in Nigeria (Yaro, 1967:210) 1973:1338.—Upatham and Sturrock, 1973:448-452.— Opuru (in the Akpada dialect of Ijaw), in Nigeria (Yaro, Bonnelly de Calventi, 1974b:35, 38, 39, figs. 6, 12.— 1967:210) Leveque, 1974:42, fig. li.—Alayo, 1974:22, pi. v: fig. 9; pi. vi: fig. 12.—Abele, 1975:56, 57.—Peck, 1975:308.— FAO names were recorded by Holthuis Hunte, 1975:66; 1977:373-376; 1978:135, 136, 139, (1980:69) as follows: Gabon shrimp (English), 144-146, 148, fig. 2; 1979b:231-241, figs. 1-5; 1979c:70.— Saltarelle gabonaise (French), and Camaron ga- Villamil and Clements, 1976:1, 4, 20, 25-28, 32, 34, 36, bones (Spanish). 37, 52, 59.—Carvacho and Carvacho, 1976:213, pi. 2: fig. vii.—Fryer, 1977:57, 58, 62, 63, 69-74, 90, 92-94, 98, Atya innocous (Herbst) 111-117, 125, figs. 2, 7, 12-14, 36-38, 40-44, 51, 55-69, 72, 73, 75, 76, 78, 83-86, 89-92, 95-112, 114-119.— FICURES ld,e, 2, 4a, 5-10, 12a,*, 22-30 Hobbs, Hobbs, and Daniel, 1977:150.—Hart, 1980:845, 847, 848.—Felgenhauer and Abele (in press). Astocus 988 Gronovius, 1764:231, pi. 17: fig. 6. Astacus Nososcopus Meuschen, 1778:86; 1781: [9].—Holthuis, Atyia occidentals.—Velez, 1967:42 [erroneous spelling]. 1966:237. Atyia innocous.—Alayo, 1974:25 [erroneous spelling].—Bon- nelly de Calventi, 1974a: 16. Cancer (Astacus) Innocous Herbst, 1792:62, pi. 28: fig. 3 [type- locality: Martinique; type: not extant].—Holthuis, REVIEW OF LITERATURE.—More 1966:237. than 100 years canc\er\ innocuus.—Latreille, 1817:37. elapsed between the appearance of Marcgrave's Atya occidentals Newport, 1847:159 [type-locality: Jamaica; (1648) account of "guaricuru" (= Atya scabra) in types: not extant].—White, 1847:74.—Gosse, 1851:85.— Brazil and the report of the second shrimp now A. Milne-Edwards, 1864:147.—Martens, 1872:135.—Gie- assigned to the genus Atya. Gronovius (1764), in bel, 1875:52.—Kingsley, 1878a:92, 93; 1878b:57.—Bate, his Zoophylacium Gronovicianum, presented a descrip- 1888:693.—Pocock, 1889:11-16, pi. 2: fig. 3, 3a; 1894:408.—Rathbun, 1897:44.—Ortmann, 1897:184.— tion of "Astacus 988," in Latin, and noted its Bouvier, 1904:137; 1905:110, 112, 113, 117-119, 121-123, occurrence in "Oceano Americano ad Martini- fig. 22; 1909:333; 1925:293, 311-314, 322, 323, 356, figs. cam." Fourteen years later, Meuschen (1778) pro- 700-702.—Hansen, 1925:140.—Allee and Torvik, posed the name "Astacus Nasoscopus" for it, 1927:67.—Boone, 1931:187-189, fig. 23.—Schmitt, which, as Holthuis (1966:237) pointed out, "is a 1935:135, 136, fig. 9.—Oliveira, 1945:179.—Villalobos, 1956:474.—Burgers, 1958:584.—Hart, 1961b:61, 63, 67, senior objective synonym of Cancer (Astacus) Inno- 72, 73, fig. 10; 1964:334.—Davant, 1963:42,44,98,100.— cous Herbst (1792), but is unavailable as Velez, 1967:42.—Straskraba, 1969:17.—Odum, 1970:H- Meuschen's publication is ruled invalid under 6.—Hume, 1975:66.—Villamil and Clements, 1976:1. Opinion 260 of the International Commission on Atya robusta A. Milne-Edwards, 1864:148, pi. 3: fig. 1 [type- Zoological Nomenclature (1954, Opin. Decl. Int. locality: New Caledonia; syntypes: MHNP 600 (<3), 1003 Comm. Zool. Nomencl., vol. 5, pt. 21, p. 267)." (6)].—Bate, 1888:693.—Ortmann, 1890:466; 1895:409.— Bouvier, 1904:137; 1905:110, 112, 116, 117, 119, 120, 122, The names proposed by both Meuschen and 128, fig. 21; 1925:292, 310-312, 314-317, 322, 323, 356, Herbst were based upon the specimen described figs. 697-699.—J. Roux, 1926b:217, 218.—Holthuis, and illustrated by Gronovius, and whereas 1966:237, 238; 1969:92.—Chace and Hobbs, 1969:58. Meuschen simply applied a name to Gronovius' Atya tenella Smith, 1871:94, 95 [type-locality: Hacienda El species, Herbst presented a translation of Gron- Polvon, Departamento Occidental, Nicaragua; syntypes: MCZ 315 (4$), PM 1785 (d\ ?)].—Kingsley, 1878a:92; ovius' description and copied his illustration. He 1878b:57.—Pocock, 1889:16.—Bouvier, 1905:121; 1925: erred, however, in noting that the locality from 312.—Oliveira, 1945:179.—Holthuis, 1966:238—Chace which the shrimp was collected was unknown (see NUMBER 364 57

above). Latreille (1817) did nothing more than the rostrum and orbital region, and summarized call attention to the species. Not until 30 years its distribution: Jamaica, Dominica, Martinique, later was it mentioned in the literature again and questionably Saint Thomas. The island of when Newport (1847) presented a brief descrip- Cuba was added within the range by him tion of specimens from Jamaica that he desig- (1909:333). Hansen (1925), in comparing the ap- nated Atya occidentalism noting that "the species pendages of crustaceans, found that the "preis- seems to be common to the West India Islands, chium" is not clearly defined in any of the pe- and appears to be that which is figured and reiopods of A. occidentalis. In his monograph of the described, but not named, by Gronovius ... ." family Atyidae, Bouvier (192,5) continued to use Obviously he was unaware of the proposed names the combination Atya occidentalis; following his of Meuschen and Herbst. The discovery of this key, he presented a synonomy including, in ad- shrimp on Jamaica by Gosse (1851:85) was re- dition to citations to Newport (1847), A. Milne- corded by him as follows: "Some interesting Crus- Edwards (1864), Von Martens (1872), Kingsley tacea are also found in Bluefields rivulet. I ob- (1878a), Pocock (1889), and Bouvier (1904,1905), tained in some numbers a new Atya which has Smith's (1871) Atya tenella and part of Ortmann's been since described by Mr. Newport under the (1895) A. scabra. A description of the species was name of occidentalism ... ." The opinion was ex- offered, affinities were discussed, and Nicaragua pressed by A. Milne-Edwards (1864) that the was added to the range given in 1905. Atya robusta characters on which A. occidentalis was established was recognized by him as a distinct species. Jean were insufficient to distinguish it from Atya scabra Roux (1926b) also considered the latter to be a (Leach, 1815), yet he described Atya robusta, be- valid species but questioned its occurrence on lieved by him to have been collected on New New Caledonia because of 150 atyids collected Caledonia. there from six localities not one was a member of One of three new members of the genus that the genus Atya (as restricted herein). Allee and had been collected on the Pacific slope of Nicar- Torvik (1927) reported the occurrence of this agua by J.A. McNiel was designated Atya tenella shrimp in Panama as did Boone (1931) who by Smith (1871). Neither White (1847), Giebel presented a long, detailed description based on (1875), nor Bate (1888) added to our knowledge six specimens from Isla Barro Colorado. Schmitt of the species, and whereas Kingsley (1878a) con- (1935) added Puerto Rico to the range of the trasted it with his A. punctata (= A. scabra), no new species, but Oliveira (1945) contributed no origi- information was offered in that publication or in nal information. another (1878b). Villalobos (1956), in describing Atya ortman- Pocock (1889) presented a detailed description nioides, pointed out its affinities with A. occidentalis. of Atya occidentalis from two localities on Dominica Burgers (1958) reported that an extract from the and pointed out the most obvious differences eyestalk of this animal, when injected into an between it and A. scabra. In his report of 1894, he "eyestalkless" fiddler crab Uca rapax, caused pig- cited two new locality records for the species on ment dispersion in the walking legs of the crab. the island of Saint Vincent. Ortmann (1890, Hart (1961b) reported the distribution of this 1895) added no new information; soon thereafter shrimp on Jamaica and noted that it is used by (1897:184), however, he listed Atya occidentalis, A. the local inhabitants as food. There the animals punctata, and A. tenella among the synonyms of A are collected by "holding baskets made of reeds scabra but again presented no original data. Rath- in the swift waters and then turning over rocks a bun (1897) also considered A. occidentalis to be a few feet upstream" (p. 73) (see Figure 8b,e herein). probable synonym of A. scabra. No new information was added by Davant (1963) Bouvier (1904) added three new localities for who suspected the occurrence of this shrimp in A. occidentalis on the island of Martinique and in Venezuela. Hart (1964) suggested that the deli- 1905 included it in his key, described it, illustrated cate texture of the carapaces of Puerto Rican 58 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY specimens of A. occidentalis was associated with the Chace and Hobbs (1969) reported the occur- small amount of calcium available in the water rence of Atya innocous in 49 stations on Dominica, from which they were collected. on nine additional West Indian islands, and from Apparently unaware of Holthuis' (1966) lucid Nicaragua to Panama. They presented a key to discussion of the synonomy of A. occidentalis with the Atya occurring in the West Indies and, for A. innocous, Velez (1967) included the species in each, included a diagnosis and notes on color, his checklist of the freshwater decapods of Puerto size, ecology, and life history. Odum (1970) and Rico under the former combination. Likewise, Chace (1972) did not add to the information Straskraba (1969) employed the same combina- contained in previous references. Bonnelly de Cal- tion in listing the freshwater crustaceans of Cuba. venti et al. (1973) and Bonnelly de Calventi The most recent references to this shrimp as A. (1974a) recorded the shrimp in the Canada Mad- occidentalis were that of Hunte (1975:66) mention- rigal, in the Dominican Republic. The latter ing Hart's (1961b) report of the species on Ja- author (1974b) presented a brief diagnosis, illus- maica and that of Villamil and Clements (1976), trations, and statements relating to color, size, who cited Schmitt's (1935) account of its presence habitat, and distribution. She found it to be less on Puerto Rico. common in the Dominican Republic than A. The delayed acceptance of the name Atya ro- scabra. Upatham and Sturrock (1973) investigated busta as a synonym of A. occidentalis (and ulti- the effectiveness of this shrimp along with several mately of A. innocous) no doubt resulted from the other animals as decoys to miracidia of Schistosoma belief that the source of the types was New mansoni. They concluded that it is unlikely that Caledonia. Even Ortmann (1890), who all too the shrimp would prevent natural transmission of frequently invoked synonomy, recognized the spe- the fluke but might "limit its severity" since some cies and assigned it to his "margaritacea-Gruppe" miracidia were found on the exoskeleton. Alayo (based on rostral characters) of the genus. The (1974) stated that A. innocous and A. scabra are identity of A. Milne-Edwards' A. robusta became common in the eastern part of Cuba, and Le- evident when Holthuis (1966) presented a con- veque (1974) reported the presence of the former vincing explanation that New Caledonia was not on Guadeloupe where he found ovigerous females likely the source of the types of either Atya robusta in May. Peck (1975) cited this shrimp as a cave or A. margaritacea. Whereas it was Holthuis' opin- inhabitant on Jamaica but did not designate a ion that the types of both came from the same specific locality. Abele (1975), in describing Atya locality in "eastern America," in view of the fact dressleri, noted the occurrence of A. innocous in that we have discovered that the types of the streams of the Atlantic Basin of Panama, pointing latter are conspecific with Smith's Atya rivalis, a out the presence of short, strong denticles on the species known only from the Pacific versant of ventral margin of the third through the fifth the Americas that shares stream habitats with A. abdominal pleura as a feature that might be used tenella (= A. innocous), which, in turn, appears to to separate the latter from the former. He was of be inseparable from A. robusta, we suggest that if the opinion that Atya tenella is distinct from A. these specimens shared a common locality it is innocous and called attention to the absence in the situated somewhere along the Pacific slope of former of "strong short denticles" on the ventral Middle America. Strengthening Holthuis' con- margins of the third through fifth pleura that are clusion that the type-locality of the two species is present in A. innocous. Differences were also sus- somewhere other than on New Caledonia is the pected by him in the "spination and form of the fact that no unquestioned records exist for mem- antennal peduncles, in the shape of the preanal bers of the genus Atya (as restricted herein) in carina and in the robustness of the legs . ..." (p. Oceania or the Indo-Pacific. Beatty (1968) re- 57). Hunte (1975) added no new data concerning corded this shrimp from several localities on Saint this shrimp. Villamil and Clements (1976) stated Croix. that in their study of the shrimp in the upper NUMBER 364 59

Espiritu Santo River they encountered only one press) provides the first recorded observations on specimen of A. innocous, that in a riffle area (p. mating behavior among members of the genus 26), and they concluded that the species appears Atya. They observed that in an aquarium, a newly to be "restricted to lower elevations," at least molted female of Atya innocous attracted (perhaps during the time of their study (p. 28). They by a pheromone) males that followed her in reported that Gifford and Cole (1970) found that swimming about the tank. When she came to rest this shrimp tends "to favor the slower flowing on a rock, a male approached, touching her with streams." "Zoea" are reported by Villamil and his antennae and third, fourth, and fifth pereio- Clements (p. 37) "to live in 15% sea water." pods, and, moving parallel to her body, repeat- Hobbs, Hobbs, and Daniel (1977) called atten- edly touched her with his extended third and tion to Peck's report of the occurrence of this fourth pereiopods. Her then climbed onto her shrimp in the cave fauna of Jamaica. cephalothorax only to be thrown off; the same The study of this species by Fryer (1977), relat- ritual was followed by other suitors. After other ing certain morphological features to the mode of prenuptial meanderings of the female and an life of the animal, is beautifully presented. This accepted partner, he moved almost directly be- report, based largely on a study of Dominican hind her, and as she moved backward, he walked members of the species, endeavors to correlate forward climbing upon her body. Shortly there- body structure with behavior, adaptations to their after, his abdomen was shifted to the side and environment, and with capture and mechanical then under hers, and, continuing in an upside- manipulation of food through the foregut. As a down position, he aligned his body with hers at result of Fryer's study, that of Hunte (1977, see which time it was assumed that the spermato- below), and Felgenhauer and Abele (in press), phore was transferred to the female. Presumably more is known of the biology of A. innocous than this was accomplished with the aid of the first of any other member of the genus. Hunte (1977), and second pleopods of the male. Some 24 hours in recounting his rearing this shrimp stated that later eggs (0.6 X 0.7 mm) were deposited on the it is the largest of the West Indian atyids, attain- pleopods of the female. The courtship and am- ing a length of about 85 mm. He found that while plexus are effectively illustrated by line drawings the first larvae did not feed, subsequent instars and photographs depicting events observed in the thrived best, in the laboratory, on wheat germ aquarium. The authors also presented excellent and Tetramin that were given them every 12 photographs, made with the aid of a scanning hours. Of the salinities and temperatures pro- electron microscope, of parts of the first and vided, 30%o and 27° C produced maximum sur- second pleopods of the male and proposed a vival. Moults occurred about four to five days hypothesis as to how the spermatophore is trans- apart, and growth was irregular. Approximately ferred to the body of the female. 80 days were required from hatching to meta- PUBLISHED ILLUSTRATIONS.—The earliest illus- morphosis to the juvenile stage. The ecological trations of this species are that of Gronovius and geographical distribution of this shrimp on (1764), a dorsolateral view of an entire animal, Jamaica was discussed by Hunte (1978). His and a redrawn, hand-colored copy of this figure study of larval development was detailed by him presented by Herbst (1792). The original descrip- the following year (1979b; see "Life History tion of Atya robusta by A. Milne-Edwards (1864) Notes" herein), and he included A. innocous in his was accompanied by a lateral view of the animal list of atyid and palaemonid shrimps of Jamaica and dorsal and lateral views of the rostrum. Po- (1979c). Hart (1980) summarized the distribution cock (1889) presented a detailed lateral view of of the freshwater atyids and palaemonids occur- the shrimp along with a drawing of the third ring in the Lesser Antilles, citing the first record pereiopod. Bouvier (1905) illustrated Atya robusta of the occurrence of A. innocous on Tobago. (fig. 21) by a dorsal view of the rostrum and a The recent study of Felgenhauer and Abele (in lateral view of the cephalic region of the carapace; 60 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY the same characters were illustrated for A. occiden- caudoventral angle of fourth and fifth pleura talis (fig. 22) along with drawings of the first subacute to acute but not produced. Sternum of pleopod of the male and female. In his mono- sixth abdominal segment more than half as long graph, Bouvier (1925) included figures similar to as wide; compressed median tubercle on sternum those published earlier but added an illustration of fifth abdominal segment small and compara- of the terminal part of the fifth pereiopod of A. tively inconspicuous. Preanal carina with com- robusta and omitted that of the first pleopod of the pressed spine overreaching posterior extremity of female of A occidentals. Boone (1931) presented a basal part. Telson 1.9 to 2.2 times as long as wide lateral view of an entire animal, and Schmitt with 6 to 8 spines in each of 2 dorsal rows. (1935) redrew Bouvier's (1905) figure of A. occi- Antennular peduncle with dorsal surface of prox- dentalis. Photographs of a Jamaican specimen in imal article usually devoid of (rarely with 1) dorsal and lateral views were presented by Hart sclerotized denticles proximal to distal row; pe- (1961b). Chace and Hobbs (1969) depicted a nultimate article 1.3 to 1.8 times as long as wide dorsolateral view of a male showing the color and dorsal surface with several to many scattered pattern (Figure 24 herein), a dorsal view of the spinules. Coxae of third and fourth pereiopods cephalic region, lateral views of the preanal car- lacking prominent anterolateral spine. Third pe- ina and third abdominal pleuron, mesial views of reiopod with merus rounded ventrally, 3 to 6 the distal part of the second pleopod and appen- times as long as high, ventromesial surface slightly dix masculina, and a view of one of its habitats to strikingly bowed, never parallel to that of on Dominica. Bonnelly de Calventi (1974b) in- corresponding podomere of other third pereiopod, cluded photographs of this shrimp in dorsal and and lateral surface bearing corneous, subsqua- lateral views and drawings of the rostrum in mous tubercles, latter frequently crowded, and, dorsal aspect, the appendices masculina and in- at least in part, linearly arranged; propodus 2.5 terna, and of the antennular peduncles. Fryer to 3.5 times as long as broad, its extensor surface (1977) presented a line drawing of a lateral view studded with strongly sclerotized spines or tuber- of an animal; others include the distal parts of cles, and flexor surface with similarly sclerotized the third through fifth pereiopods, details of the spines or tubercles where if any arranged in row structure, musculature, and setae of the first pe- those comprising latter never contiguous or over- reiopod, and photographs of a feeding animal: lapping; spines and/or tubercles flanked distally filtering, sweeping, and scraping (some showing and to the sides by semicircular, often conspicu- color pattern). In addition, drawings of the seta- ous, patches of plumose setae; dactyl freely mov- tion of the gnathal appendages and detailed ren- able and bearing 2 oblique rows of scale- or ditions of the anatomy of the foregut are included. spikelike denticles on flexor surface. The most recent illustrations pertaining to this MALE (Mannet's Gutter, Dominica, WI).— shrimp are those of the larvae and their append- Rostrum (Figure 22a,d) with margins strongly ages presented by Hunte (1979b) and of courtship contracted at base of acumen (often forming and details of the first and second pleopods of the angles); apex of acumen slightly overreaching male by Felgenhauer and Abele (in press). proximal podomere of antennular peduncle; dor- DIAGNOSIS.—Cephalic region of carapace not sal median carina gently curved, not excavate conspicuously sculptured, glabrous; antennal and dorsally (not dipping below level of lateral cari- pterygostomian spines prominent. Rostrum with nae posterior to acumen), and ending preapically margins suddenly contracted at base of acumen on acumen; ventral carina with 3 small teeth near forming subangular bends; angles never pro- apex of acumen; ocellar beak hidden in lateral duced acutely anteriorly. Ventral margin of third aspect by eyes, reaching level of about midlength through fifth abdominal pleura usually, except in of stylocerite, its cephalic border vertical, and Pacific watershed, provided with rows of sclero- rounded dorsal margin embraced by sides of ven- tized denticles (such never present on second); tral rostral groove. Antennal and pterygostomian NUMBER 364 61

FIGURE 22.—Atya imocous (all from males from Mannct's Gutter, Dominica): a, dorsal view of cephalic region; b, dorsal view of antennular peduncle; c, mesial view of appendices masculina and interna; d, lateral view of carapace; t J, lateral view of second through fifth abdominal pleura; g, flexor surface of distal part of fourth pereiopod; A, flexor surface of dactyl of fourth pereiopod; t, flexor surface of distal part of third pereiopod; jjc, flexor surface of distal part of fifth pereiopod; /, lateral view of third pereiopod; m, lateral view of preanal carina; n, ventral view of basal podomeres of third pereiopod; o, dorsal view of telson. (Scales marked in 1 mm increments.) 62 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY spines strong; no spines present between them. podomere; penultimate segment about 1.1 times Surface of carapace with crowded punctations as long as ultimate. supporting very short, erect, fine setae dorsally, First pereiopod reaching distal end of anten- latter nowhere conspicuous; devoid of ridges and nular peduncle; second overreaching first by spines other than those just mentioned. about one-third length of fingers; terminal brush Pleura of first 4 abdominal segments (Figure of both appendages containing setae with scrap- 22;) with rounded to subacute posteroventral ex- ing denticles. Third pereiopod (Figure 22t,/,n) tremities; corresponding parts of fifth acute but without lateral distoventral spine on merus and not produced in spine. Third through fifth pleura carpus, ventral spine absent from merus, and with fine corneous denticles on ventral margin carpus lacking distolateral spines; when append- and fifth with moderately conspicuous fringe of age extended anteriorly, overreaching antennular plumose setae. Fourth abdominal tergum almost peduncle by dactyl, propodus, and half length of 1.3 times as long as fifth (latter subequal to sixth) carpus. Merus with ventromesial margin bowed, and 1.1 times as long as telson. Sternum of fifth almost 3 times as long as high, 2.2 times as long abdominal segment with small compressed me- as carpus, almost 2.4 times as long as propodus; dian tubercle (Figure Id). Sternum of sixth ab- propodus 2.8 times as long as wide, 0.9 as long as dominal segment about 0.8 as long as broad. Free carpus; distoventral margin of coxa entire (evenly part of preanal carina (Figure 22m) spiniform, rounded), and mesial caudoventral prominence curved, and slightly overreaching angle of basal virtually absent, distal ventrolateral spine also part of sclerite. Telson (Figure 22o) about 1.4 lacking. Lateral, dorsal, and ventral surfaces of times as long as broad, its dorsal surface bearing merus studded with linear series of large tubercles, paired concave rows of 6 corneous denticles and most bearing cornified discs apically and flanked posteromedian tubercle, latter overhanging cau- by distal arc of setae; conspicuous tufts of long dal margin. plumose setae present ventrolaterally; mesial ex- Proximal podomere of antennule (Figure 22b) tremity of podomere produced in rounded lobe with stylocerite reaching distal fourth or fifth of at level of mesial articular condyle of carpus. segment; dorsal surface with linear cluster of setae Latter strongly tuberculate, and tubercles tipped but lacking corneous spinules; distal margin bear- with corneous discs and flanked distally by arc of ing row of 9 corneous spinules; penultimate seg- 3 or more setae; setae long and forming conspic- ment of peduncle about 1.4 times as long as wide uous tufts ventrolaterally. Propodus also strongly and provided with 23 (right) or 24 (left) scattered tuberculate with corneous elements (discs) pro- corneous spinules on dorsal surface and row of 11 duced in conspicuous corneous spines; tubercles (right) or 10 (left) on distal margin; ultimate on flexor surface somewhat irregularly arranged podomere with 10 (right) or 9 (left) spinules on in proximal half of podomere, tending toward dorsal surface, row of 9 (right) or 8 (left) at base biserial arrangement in distal half; lateral surface of lateral flagellum, and row of 6 at base of mesial bearing tufts of plumose setae. Dactyl movable, flagellum. Antenna with ventrolateral spine on its flexor surface with 2 irregular, oblique rows of basis reaching as far anteriorly as pterygostomian denticles flanked distally by paired clusters of spine but not quite so far as stylocerite; lateral setae. spine on scaphocerite strong, reaching about mid- Fourth pereiopod with dactyl reaching distal length of ultimate podomere of antennular pe- end of proximal third of carpus of third pereio- duncle; lamella far surpassing latter; flagellum pod; merus about 1.8 times as long as carpus; extending slightly beyond midlength of telson. latter subequal in length to propodus. Fifth pe- Third maxilliped overreaching antennular pe- reiopod reaching distal extremity of carpus of duncle by two-thirds of distal podomere of en- fourth pereiopod; merus 1.3 times as long as dopod; tip of exopod attaining base of ultimate carpus, latter approximately 0.8 as long as pro- NUMBER 364 63

FIGURE 23.—Atya innocous (a, male syntypc of A. lennella; b-dji-j,l, male from Cerro Azul area, Panama; e-gjc, male from Rio Guabas, Panama): a, dorsal view of cephalic region; b, dorsal view of antennular peduncle; c, mesial view of appendices masculina and interna; d, lateral view of cephalic region; e, flexor surface of distal part of fourth pereiopod; /, flexor surface of distal part of fifth pereiopod; g, flexor surface of distal part of third pereiopod; A, lateral view of second through fifth abdominal pleura; t, lateral view of third pereiopod; j, lateral view of preanal carina; k, flexor surface of distal part of fifth pereiopod; /, dorsal view of telson. (Scales marked in 1 mm increments.) 64 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY podus. Ornamentation of merus, carpus, and pro- gray at base fading to cream distally. Basal podomeres (coxa podus of fourth pereiopod consisting of distal through ischium) of third leg cream with irregular dark ventrolateral spine, 2 more proximal ventral ones brown splotches; merus light basally, becoming dark brown distally and bearing very dark brown tubercles, few with on merus, and distal ventrolateral and 3 distola- corneous tips; carpus dark brown with light tan band at teral spines on carpus; ventrolateral surface of midlength, propodus light tan in proximal fourth and dark merus, carpus, and propodus with conspicuous brown in distal three-fourths, tubercles on carpus and pro- band of plumose setae. Ornamentation of corre- podus dark at base but with corneous (yellow) tips; dactyl sponding podomeres of fifth pereiopod similar mostly corneous. Fourth and fifth pereiopods with basal podomeres as in third pereiopod; merus with flexor portion, but with 3 ventral spines on merus and 2 disto- proximal and distal ends dark brown, remainder tan; carpus lateral spines on carpus. with proximal extensor surface tan, otherwise dark brown; Diaresis of lateral ramus of uropod flanked propodus and dactyl as on third pereiopod. proximally by row of 23 (right) or 22 (left) artic- In most young specimens, light dorsomedian stripe ex- ulated, corneous denticles, and fixed spine at tending from tip of rostrum almost to, or to, distal [sic, = lateral end of row. posterior] margin of telson; on cephalothorax, stripe of uni- form width; on first abdominal somite, expanded in posterior COLOR NOTES.—The most complete account of half; in succeeding three somites, narrow anteriorly and the color of Atya innocous is that of Chace and broadening posteriorly; in fifth somite narrow and of uni- Hobbs (1969:58-60). form width; in sixth, essentially similar to that in second through fourth; and on telson, narrower posteriorly. In older Two phases, green and brown. individuals, dark pigment forming variable, mostly bilat- Brown Phase: Ground color of cephalothorax dark brown erally symmetrical, patterns along lateral margins of dorso- (brownish black chromatophores forming reticulate pattern median stripe; with increasing age patterns coalescing and over tan). Dorsum of carapace with dark brown longitudinal infringing on stripe to extent that in individuals of inter- stripe extending from base of rostrum to posterior margin, mediate size, stripe usually narrower, irregular, and inter- becoming broader posteriorly. Stripe continuing onto ab- rupted, and largest individuals usually without trace of domen but broken by very narrow transverse tan bands stripe. across posterior margins of anterior five terga; sixth tergum tan anteriorly and dark brown posteriorly; broadest portion Green Phase: Pattern essentially identical; color, however, of stripe on first and second somites, in both becoming ranges from pale bluish green to greenish black. narrower posteriorly. Lateral surface of carapace straw The pattern of a male with a carapace length brown with several obliquely directed dark brown lines, posterior and ventral ones directed posteroventrally; poster- of 25.8 mm is reproduced from Chace and Hobbs olateral area with lateral and ventrolateral pale tan spots in Figure 24. Bonnelly de Calventi (1974b:39) outlined above and below by aforementioned pairs of dark observed that the median dorsal stripe "desapa- brown lines. Ventrolateral portion of abdominal terga and rece con los cambios de luz del medio ambiente." pleura mottled straw brown with short dark brown lines (in anterior somites) or spots (in posterior somites) adjacent to Fryer (1977:70) confirmed the observations of articular knobs at posterior bases of pleura: second pleuron Chace and Hobbs that the color pattern changes with prominent pale tan spot at base and succeeding three with age and added that "many individuals, pleura with similar, progressively smaller ones, all forming especially larger ones appear to be almost black." row with lateral spot on carapace; each spot on pleura with Several of the photographs therein show the color dark border. Ventral and posterior margins of pleura very light with dark submarginal line. patterns mentioned above. The color pattern of specimens from El Valle, Antennular peduncle straw brown with dark brown rings, antennal peduncle straw brown with dark brown markings Panama, differs in no conspicuous way from that just proximal to, and on lateral margin of, antennal scale; described above. The most obvious difference flagella dark brown to tan. Third maxillipeds pale, translu- between the color photographs lent to us by Bruce cent, with narrow black lateral margins on more distal E. Felgenhauer and the pattern illustrated in podomeres. First two pereiopods translucent to straw brown Figure 24 is the presence of a broad, median, with dark brown lateral line on merus; carpus straw brown with orange spot on articular surface; bases of two distal longitudinal, light stripe extending from the apex podomeres orange, followed by bluish cream throughout of the rostrum onto the telson; flanking the stripe most of their lengths, with subterminal narrow, vivid orange is a pair of irregular, narrow, very dark brown to band and terminal white one bearing setal tufts; setae dark blackish lines. Also, the more ventral of the two NUMBER 364 65

FIGURE 24.—Lateral view of Atya innocous (from Chacc and Hobbs, 1969). light spots on the posterolateral surface of the had carapace lengths of 5.0 to 33.7 mm; the range carapace of the Dominican specimen is lacking in in 246 females was 2.5 to 20.6 mm, including 80 that from Panama, and the bands on the antennal ovigerous ones, which measured 8.8 to 20.6 mm. peduncle of the latter are not so clearly defined, The carapace length of 215 juveniles ranged from rather the two distalmost podomeres are dark 1.1 to 5.0 mm. Fryer (1977:69) noted that the brown dorsally with only faint suggestions of the maximum total length of members of this species bands laterally. is about 12.2 cm. Of the specimens measured by Hunte (1979b) recorded the positions of chro- us, the largest is a male from Dominica, having matophores in the first larval stage of A. innocous, a carapace length of 40.9 mm. The ovigerous noting that on a "virtually transparent" animal, females ranged from 7.9 to 24.9 nr reddish chromatophores were present as follows: DISTRIBUTION AND SPECIMENS EXAMINED.—This one on the antennular peduncle, one on the car- shrimp, described from Jamaica, has a range in apace posterior to each eye, three on the first the Americas almost as great as that of Atya scabra abdominal segment, and at least one on each of and, where it occurs, appears to be far more the following segments. Heavy pigmentation oc- abundant. It has been reported on the Pacific curred at the junction of the carapace and ab- slope as A. tenella, from Nicaragua to Panama; in domen and at that between the latter and the the Atlantic watershed, it ranges from Nicaragua telson. Four chromatophores were found on the to Panama, and in the West Indies it has been telson. recorded from the Greater Antilles and from the SIZE.—Although he provided only one mea- Virgin Islands southward to Curacao. Chace and surement of an adult, about 85 mm total length, Hobbs (1969) reported its presence on the island Hunte (1979b:231) stated that Atya innocous is the of Dominica in some 49 localities ranging from largest atyid shrimp occurring in the West Indies. sea level to 925 m above sea level (Figures 25-28). Of the almost 700 specimens from Dominica Records for the known localities are listed be- examined by Chace and Hobbs (1969), 232 males low. Collections that we have examined are 66 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

*USNM, trib to near Elysium at 76 m (Hart, 1961b:72), 16* (10.5), 2$ (6.5, 13.9), 1 ovig 9 (10.4), 8 Apr 1959, CWH, G. Thomas. (3) Drivers River at 362 m (Hunte, 1978:140). (4) River at 300 m (Hunte, 1978:140). (5) Buff Bay River at 910 m (Hunte, 1978:140). (6) fMCZ, brook at Spring Bank 4.2 km W of Port Antonio, 16* (23.4), 1 ovig 9 (16.2), 28 Mar 1906, A.E. Wright. (7) fMCZ, Brook Port 3.4 km W of Port Antonio, 56* (8.5-15.5), 27 Mar 1906, AEW. (8) fUSNM, Mabess River, 26* (approx 31.0, 36.0), 26 Jul 1926, D.S. Johnson. (9) fUSNM, Falls, 36* (10.5-13.3), 1$ (13.8), 1 ovig 9 (16.6), 11 Jul 1948. (10) fLGA, John Crow Mts at about 600 m, 1 ovig 9 (23.9). Saint Andrew Parish—{\\) *USNM, Mammee River near Maryland at 380 m (Hart, 1961b: 72), FIGURE 25.—Distribution of Atya innocous (circled numerals 16* (15.2), 5$ (6.2-13.5), 4 Jun 1959, CWH, TF. = number of localities; see Figures 26-28). (12) at 1050 m (Hunte, 1978:140). (13) at 575 m (Hunte, 1978:140). (14) marked with an asterisk if they have been previ- fUSNM, near Cinchona at 1200 ously reported and with a dagger if they are m, 26* (32.0, 36.1), 3 ovig? (23.6-24.0), 1910, E.A. reported herein for the first time. Numbers follow- Andrews. (15) fUSNM, Yallahs River near Cin- ing the specimens listed are measurements, in chona, 16* (33.0), 1 ovig 9 (24.9), 10 Jun 1910, mm, of the carapace length or, if followed by EAA. (16) fUSNM, Upper Ginger River, Tweed- "t.L," total length. Some listings lack dates and/ or collectors; these could not be determined. CUBA (all localities in Provincia de Oriente): No specific locality, (Bouvier, 1909:333). (1) *MHNP, eastern extremity of Cuba (Bouvier, 1925:313), 1 ovig 9 (15.6), 1910. (2) fUSNM, Rio Santa Maria de Loreto, Ramon de las Yaguas at about 480 m, 1$ (11.6), CJ. Ramsden. (3) fUSNM, Arroyo de la "Alcachofa" (probably Alcahuete) at about 360 m, 2$ (8.5, 14.4), 3 Mar 1919, CJR. (4) fUSNM, Rio Los Hondones, Guantanamo, at about 490 m, 1$ (25.2), CJR. (5) fUSNM, Arroyo Los Machitos, El Corojo de San Carlos, Guantanamo, 16* (26.2), CJR. (6) fMCZ, Pico Turquino, 16* (24.8), 10-19 Jun 1936, J. Acuna. JAMAICA: No specific locality (Newport, 1847:159); *USNM, 16* (22.2), 1$ (19.5), 1-11 Mar 1884, Str. Albatross. Manchester Parish—(I) *USNM, One Eye River S of Auchtembeddie at 230 m (Hart, 1961b:72), 2$ (18.0, 22.0), 13 Jan FIGURE 26.—Distribution of Atya innocous in Costa Rica and 1960, CWH, T. Farr. —(2) Panama. NUMBER 364 67 side at 600 m, 36 (29.0-32.2), 2$ (16.4, 21.0), 2 Laughland's at 15 m (Hunte, ovig 9 (20.9, 21.0), 13 Apr 1903, W.R. Maxon. 1978:140). (20) Laughland's Great River at 304 (17) fUSNM, Clyde River at Chestervale, \6 m (Hunte, 1978:140). (21) at 152 m (30.3), 1$ (22.5), 30 Jun 1948. — (Hunte, 1978:140). (22) fBM, Ocho Rios, 29 (18) *USNM, Great River at Llandovery at 30 m (10.9, 15.4), 16 Jan 1937, I. Sanderson. Saint (Hart, 1961b:72), 7(5 (11.1-17.4), 5? (6.0-12.1), 6 Catherine Parish—(23) Rio D'Oro at William's ovig 9 (9.0-12.6), 10 Apr 1959, CWH, GT. (19) Field at 150 m (Hart, 1961b: 72). Saint James

FIGURE 27.—Distribution of Atya innocous in the Greater Antilles. 68 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

63' Jan 1960, CWH. (26) Rio Nuevo at 454 m (Hunte, 1978:140). (27) fUSNM, Rock Spring Cave, Pear Tree Grove, 4(5 (15.7-31.7), 7$ (16.2-18.7), 1 ovig? (20.8), 21 Aug 1974, S. Peck. Saint Thomas Parish— (28) Bugaboo River 0.3 km below Corn Puss Gap at 600 m (Hart, 1961b:73). (29) 4.3 km NW of Trinity Ville at 450 m (Hart, 1961b:73). (30) *USNM, Banana River 1.7 km N of Richmond Vale at 450 m (Hart, 1961b:73), 29d (9.8-22.5), 5$ (9.1-13.6), 27 ovig 9 (9.9-18.0), 6 Jan 1960, CWH, R. Ben- gry. (31) Morant River at 300 m (Hunte, 1978:140). (32) Bugaboo River at 543 m (Hunte, 1978:40). (33) Yallahs River at Ramble Bridge at about 82 m (Hart, 1961b:73). — (34) below dam near Martha Brae at 15 m (Hart, 1961b:73). (35) fRNHL, Martha Brae River S of Bunkershill, 26 (15.7, 22.8), 17$ (7.8-14.4), 10 Jan 1960, CWH. HISPANIOLA: Haiti—no specific locality (King- sley, 1878a:92); *MCZ, 36 (11.3-18.6), 8 Sep 1858, D.F. Weinland. (1) fBM, Thorlands, 86 (7.8-11.6), 3$ (8.0-9.7), 1 ovig $ (11.8), 17 May 1937, IS. (2) fUSNM, Port au Prince, 1$ (in- jured), 20 May 1930, W.M. Parrish. Dominican Republic—no specific locality, fUSNM, 36 (17.0-25.1), 2$ (13.4, 18.1), 4 ovig 9 (13.9-17.4), 1878, W. Gabb. (3) Canada Madrigal, Distrito Nacional (Bonnelly de Calventi et al., 1973: 1338); 66 (13.8-25.4), Bonnelly de Calventi, 1974a: 16. PUERTO RICO: (1) Luquillo Forest (Schmitt, 1935:136). (2) fRNHL, Rio Guanica, 5 juv (2.0-2.1), 15 Sep 1965, P.W. Hummelinck. (3) fUSNM, Rio Manati between Corozal and Or- ocovis, 26 (9.5, 10.3), 1$ (9.8), 23 Nov 1954. (4) fUSNM, trib to Rio Mameyes at Rte 112, 2.7 FIGURE 28.—Distribution of Atya imocous in the Lesser An- and 1.8 km N of El Yunque summit, \6 (21.3), 8 tilles (circled numerals = number of localities). Jun 1953. (5) fUSNM, Rio Maracayo, pool below fish hatchery, 26 (36.8, 37.0), 1 Feb 1971, D.S. Erdman. Parish—(24) fUSNM, near Tan- SAINT CROIX: Beatty (1968:263) cited the follow- nery, 16 (28.3), 10 Jul 1910, EAA (?). Saint Mary ing: "Upper Love, Fairplain, Caledonia, Crique, Parish—(25) *USNM, trib to Trunnels River 6.8 La Grange, streams, Mt. Welcome Swamp." km S of Richmond at 230 m (Hart, 1961b:72), Chace and Hobbs, (1969:60) cited no specific 26 (10.5, 13.2), 2 ovig 9 (8.4, 15.4), 7 Apr 1959, locality; *USNM, 16 (20.4), H.A. Beatty. (1) CWH; 26 (8.5, 12.2), 1$ (12.3), 1 ovig $ (16.6), 7 fUSNM, Caledonia Street, 16* (8.4), 29 (11.2, NUMBER 364 69

17.0), HAB. (2) fRNHL, Canaan Stream, 46 HHH III, HHH; 56 (9.4-22.7), 39 (13.0-15.2), 1 (7.1-15.5), 59 (7.7-10.4), 10 Jun 1955, PWH. ovig 9 (16.7), 10 Nov 1964, PJ. Spangler. (12) SAINT THOMAS: No specific locality (Bouvier, Mill Race to Rosalie River at 15 m. (13) fUSNM, 1905:119). trib to Fond Figues River, alt 410 m, 16* (10.2), MONTSERRAT: NO specific locality (Chace and 29 (8.3-11.3), 4 ovig 9 (12.4-16.1), 62 juv Hobbs, 1969:60); *USNM, spring, alt about 450 (5.0-7.0), 6 Mar 1966, HHH. Saint George Parish— m, 16* (8.2), 1$ (8.3), 1894, Hubbard and (14) *BM, Laudat (Pocock, 1889:11), alt 300 m, Schwartz. 46 (26.1-40.9), J.A. Remage. (15) *USNM, GUADELOUPE: No specific locality (Chace and stream at Fond Baron Estate 1 mi E of Loubiere, Hobbs, 1969:60); *USNM, stream, 16* (11.4), 1$ alt 135 m, 16 (11.7-22.8), 49 (12.2-13.8), 1 juv (14.2), 1 ovig 9 (12.7), Mar 1937, HAB. (1) (6.2), 21 Feb 1964, HHH III, HHH. (16) fMHNP, second chute du Carbet, Basse Terre, *USNM, trib to Roseau River below Trafalgar alt 600 m, 46 (16.8-22.0), 1$ (11.9), 14 Jun 1978; Falls, alt 300 m, 66 (6.0-34.2), 19 (21.1), 1 ovig 25<5 (9.1-25.2), 69 (8.6-11.5), 3 ovig $ (11.9-14.6), 9 (20.5), 5 juv (no more than 5.0), 26 Mar 1964, 24 Jun 1978. (2) fMHNP, Riviere Belle Eru CWH, HHH. (17) *USNM, pool below Trafalgar (Grand Etard), alt 180 m, Basse Terre, 46 Falls, alt 300 m, 16 (21.5), 1 ovig 9 (14.3), 8 Jan (14.1-23.0), 1$ (15.4), 1 ovig? (13.6). 1964, HHH III; 26* (25.9,32.5), 25 Oct 1964, PJS. DOMINICA: (Except for localities 2-6, 12, 20-24, (18) *USNM, stream near Freshwater Lake, trib 37, 39-44, 58, and 59, all of those from the island to Rosalie River, alt 760 m, 16 (21.3), 23 Feb were included in Chace and Hobbs, 1969:9-12.) 1964, purchased by D.L. Bray. Saint John Parish— Unknown locality, *USNM, 86* (21.2-32.2), 4 (19) *USNM, Hermitage River N of Portsmouth, ovig 9 (15.4-17.2), 17 Jul 1877, F.A. Ober. Saint alt 90 m, 13c? (8.0-18.3), 119 (5.5-17.6), 21 juv Andrew Parish—(I) *USNM, trib of Pagua River (about 5.0), 5 Mar 1964, HHH. (20) Headwaters 3.4 km N of Deux Branches, alt 106 m, 36 of Picard River, alt 450 m. (21) Lamonthe River, (5.5-10.3), 4 Mar 1966, HHH. (2) Trib to Ka- alt 121 m. (22) Cario River, alt less than 8 m. siobna River, alt 91 m. (3) Toulaman River, alt (23) Barry River to Indian River, alt about 3 m. 15 m. (4) Hodges River, alt 7.5 m. (5) fUSNM, (24) fUSNM, headwaters of Espagnole River on trib to Melville Hall River 2.6 km N of airport, slope of Morne Diablotin, alt 450 m, 16* (17.2), alt 91 m, 19 (10.5), 1 ovig 9 (14.2), 3 Feb 1964, 69 (10.3-17.3), 26 Jan 1964, HHH. Saint Joseph HHH III, HHH. Saint David Parish—(6) *USNM, Parish—(25) Layou River (Pocock, 1889:11). (26) trib to Fond Figues River, to Castle Bruce River, *USNM, Mannet's Gutter near mouth, alt 15 m, alt 410 m, 16* (10.2), 29 (8.3-11.3), 4 ovig 9 56* (15.1-26.6), 79 (11.4-17.5), 38 juv (3.0-5.0), (12.4-16.1), 62 juv (5.0-7.0), 6 Mar 1966, HHH. 25 Jan 1964, HHH; 16* (10.0), 19 (14.9), 1 juv (7) *USNM, Fond Figues River at 105 m, 26 (5.2), 19 Feb 1966, HHH; 26* (18.0, 20.1), 29 (12.3, 13.5), 1 ovig 9 (14.5), 2 juv (5.0, 5.5), 22 (12.4, 16.1), 21 Feb 1966, HHH; 146* (14.0-29.2), Mar 1964, HHH. (8) *USNM, trib to Castle 99 (13.2-16.5), 7 ovig 9 (14.4-19.0), 2 Feb 1967, Bruce River W of Raymond Stone River, alt 76 C. Rhyne. (27) *USNM, Mannet's Gutter at m, 36 (8.3-14.4), 49 (8.4-12.3), 22 Mar 1964, upper bridge, alt 21 m, 96* (15.3-29.6), 59 CWH, HHH. (9) *USNM, trib to N Branch of (13.0-19.9), 4 ovig 9 (14.1-21.1), 29 Jan 1964, Ravine Deux Dleau, alt 180 m, 14c? (7.0-21.3), HHH; 36* (14.7-22.8), 29 (10.2, 17.8), 3 ovig 9 119 (6.9-12.0), 3 ovig9 (9.3-12.1), 3juv (5.0-6.1), (17.5-20.0), 7 juv (4.7-7.0), 17 Mar 1966, HHH; 14 Feb 1964, HHH III, HHH. (10) *USNM, N 16* (23.0), 19 (21.5), 1 ovig 9 (17.6), 8 Aug 1965, Branch of Deux Dleau, alt 242 m, 16* (11.1), 19 D.M. Anderson; 16* (18.3), 1 Sep 1965, DMA; 26* (12.5), 29 Jan 1964, HHH III, HHH; 46 (16.1, 31.3), 19 (18.4), 4 juv (about 5.0), 7 Mar (6.8-18.7), 19(8.2), 14 Feb 1964, HHH III, HHH. 1966, R.B. Manning, HHH. (28) *USNM, trib (11) *USNM, trib to Rosalie River near Bori of Layou River across from Clarke Hall, alt 15 Lake, alt 920 m, \6 (18.9), 23 Mar 1964, CWH, m, 76* (12.0-25.5), 29 (16.5-17.1), 15 juv (5.0), 7 70 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Fcb 1964, HHH III, HHH; 26 (10.2, 17.8), 2$ (13.0), 1 ovig 9 (9.4), 3 Sep 1965, AL; 26 (6.0, (9.5, 14.4), 13 Mar 1964, HHH; 36 (7.2-19.8), 2$ 8.8), 1 ovig 9 (10.0), 15 Mar 1966, HHH. Saint (17.2, 17.3), 29 juv (4.0-5.0), 22 Fcb 1966, HHH. Paul Parish—(51) *USNM, headwater trib of Bel- (29) *USNM, trib to Layou River at Cassada fast River, Pont Casse, alt 600 m, 1<5 (20.2), 19 Gardens, alt 150 m, 36* (13.2-24.3), 49 (15.9), 1 juv (7.6), 16 Feb 1964, HHH III, HHH. (10.6-16.2), 5 juv (about 5.0), 19 Feb 1964, HHH; (52) *USNM, trib to Belfast River at Sylvania, 16 (21.2), 4 ovig 9 (15.0-17.2), 11 juv (about 5.0), alt 575 m, 36 (8.9-17.5), 49 (10.0-17.2), 28 Jan 3 Mar 1964, H. Robinson, HHH. (30) *USNM, 1964, R. Zusi, HHH III. (53) *USNM, Check Dleau Morne Laurent to Layou River, alt 210 m, Hall River at Springfield, alt 350 m, 106* 26 (20.8, 28.0), 2$ (18.4, 25.9), 2 ovig 9 (13.1, (7.9-19.3), 129 (7.6-19.8), 4 juv (about 5.0), 1 18.5), 29 Feb 1964, HHH III, HHH. (31) Feb 1964, HHH III, HHH. (54) *USNM, trib to *USNM, trib to Layou River just N of Dleau Check Hall River, alt 460 m, 11<5 (7.2-23.2), 149 Manioc, alt 210 m, 86 (14.6-27.5), 1? (16.5), 4 (12.4-16.2), 2 ovig 9 (10.7-12.5), 1 juv (6.0), 1 ovig 9 (12.1-16.4), 20 Fcb 1964, HHH III, HHH. Feb 1964, R. Patrick, HHH. (55) Belfast River, (32) *USNM, Warner River 3.4 km N of Pont alt 23 m. (56) Mouth of Belfast River at sea level. Casse, alt 410 m, 1 ovig 9 (12.3), 4 Mar 1966, (57) Mouth of Check Hall River at less than 2 m. HHH. (33) *USNM, trib of Laurent River E of Location Uncertain—(58) fUSNM, Du Blanc River, Pont Casse, alt 485 m, 66 (7.2-12.2), 4? the Jungle Biol Sta, 29 (9.9, 14.0), 3-19 Mar (8.2-18.3), 8 ovig 9 (11.0-13.1), 4 Mar 1966, 1929, P.G. Howes. (59) fBM, Emerald Pool, be- HHH. (34) Layou River at lowest riffle, approx- neath waterfall, 19 (15.1), 1 juv (5.7), 2 Aug 1977, imately 2 to 3 m. (35) North bank of Layou River Michael New. about 30 m above mouth at sea level. (36) Batali MARTINIQUE: No specific locality (Gronovius, River near mouth at sea level. (37) Trib to Layou 1764:231). (1) *MHNP, "Rousseau" (Bouvier, River just N of Dleau Manioc. (38) Mouth of 1904:137), 76 (15.5-30.5), 59 (17.6-19.0), 1 ovig Layou River to 200 m upstream on S bank, sea 9 (15.1). (2) *MHNP, "Bellanger" (Bouvier, level. (39) Macoucheri River in vicinity of bridge, 1904:137), 76* (21.3-27.4), 2 ovig 9 (18.9, 19.6). alt 1 to 3 m. (40) Trib of Macoucheri River, mill (3) ChafTanjon (Bouvier, 1904:137), (4) fRNHL, race, alt about 2 m. (41) Trib to Layou River just Dumauze, 26 (15.1, 22.0), 13 Jun 1955, R. Pin- S of River D'Or, alt 212 m. (42) Mannet's Gutter, chon. Clarke Hall Estate, alt 106 m. (43) Ravine Neiba SAINT LUCIA: (1) fBM, Marc stream, 16* (14.0), to Layou River, alt 76 m. (44) fUSNM River 1 ovig 9 (16.3), 15 juv (3.2-6.0), 18 May 1971, G. D'Or, to Layou River, alt 230 m, 16 (30.2), 1 ovig Barnish. 9 (16.4), 29 Feb 1964, HHH HI, HHH. Saint SAINT VINCENT: No specific locality, *RNHL, Patrick Parish—(45) *USNM, Ravine Cacao, alt 36 (18.7-31.8), 1 ovig 9 (17.0), 1975, S.P. Meyers. 120 m, 1

1967, PWH. (2) fRNHL, Conset River at St 1 juv 6*, Jul 1868, J.A. McNiel. John, 96* (6.3-16.5), 4$ (6.3-9.0), 6 ovig $ COSTA RICA: (1) fUSNM, Rio de los Platanales, (9.3-13.4), 26juv (3.6-6.1), 7 Jul 1967, PWH. (3) Golfo Dulce, 16* (13.8), 2$ (15.0, 17.4), Apr 1896, fRNHL, Joe's River at Frizers, W of Bathsheba, H. Pittier. (2) fUSNM, Rio Barranca, 1$ (17.7), 136* (6.2-11.6), 7$ (7.4-16.7), 64juv (2.9-6.0), 16 4 Aug 1928, M. Caleris. Feb 1964, PWH. PANAMA: Provincia de Bocas del Toro—(1) GRENADA: (1) fRNHL, Titivee, Victoria, 1 juv fUSNM, Rio Sixaola and creek near Finca Calif, (4.1), 16 Jan 1958, PWH. (2) fRNHL, Beaulieu, pumping station, 2 juv (2.8, 3.0), 2 Sep 1962, H. examined but sex and c.l. not recorded, 3 Oct Loftin. Zona del Canal—(2) *USNM, Shannon 1965, J.R. Groome. (3) fRNHL, La Sagesse Creek on Isla Barro Colorado (Allee and Torvik, River, Coats Gap, St. Davids, 16* (15.9), 1 Jan 1927:67 for A. occidentals) 1$ (19.4), Mar 1924, 1966, Justin Francis, JRG. (4) fRNHL, St. Marks W.C. Allee. (3) fMCZ, "Limones" (= Ei Li- River, 56* (16.8-21.5), 3$ (11.0-13.6), 55 juv mon?), 1 ovig $ (about 11.0), 8 Jun 1929. (4) (1.0-5.3), 25 Jan 1966, C.A.O. Philips, JRG. (5) *USNM, Isla Barro Colorado, 16* (18.3), 26 Jun fRNHL, Irwin's River, Z.O. van Santenos, 40 1969, L.G. Abele. (5) fUSNM, Pedro Miguel juv (2.2-4.0), 7 Jul 1967, PWH. Locks, 2 juv (1.2, 1.9), 3 Feb 1969, LGA. (6) TOBAGO: No specific locality, fBM, 1 ovig $ fUSNM, Pedro Miguel, Madden Forest, Green (13.4), P.L. Guppy. (1) *USNM, Doctor's River, Memorial, 16* (17.5), 23 Jul 1969, LGA; LGA, 46 in small dammed pond near Speyside, 11°18'N, (11.0-20.7), 4$ (15.6-17.5), 1 Mar 1973, LGA, fl^m (Hart, 1980:847), 46 (17.0-29.0), 9 Apr M.H. Robison, F. Quieros; LGA, 46* (16.0-18.1), 1978, F.D. Martin et al. (2) fBM, Mt. Irvine 2$ (14.5, 16.8), 4 Apr 1973; LGA, MHR. Provincia River SW of Plymouth, 56 (17.2-24.0), A.K. de Chiriqui—(7) fUSNM, trib to Rio Tabasara 22 Totten. (3) fRNHL, Frenchman's River at Spey- km W of El Maria on Sona-Remedios Rd, 36* side, 36 (8.0-11.9), 7$ (6.0-12.8), 3 juv (6.3-6.5). (13.1-21.3), 11 Nov 1961, HL. (8) fUSNM, creek (4) fUSNM, cascading stream on Roxborough, 17 km W of David, 36* (10.6-18.3), 1$ (18.0), 2 Parlatuvier Road, vie. Bloody Bay, 16* (23.3), 31 Dec 1961, HL. (9) fUSNM, 8.5 km W of David Aug 1972, R.G. Tuck, Jr., W. Flowers; 16* (27.6), on Carretera Interamericana, 26* (13.9, 14.4), 1 4 Sep 1972, RGT. ovig $ (16.1), 12 Dec 1961, HL. Provincia de Code— TRINIDAD: (1) fBM, Mt. Aripo, 26 (8.3, 11.1), (10) fMCZ, Rio Las Lajas, 36* (7.3-11.0), 22 Oct IS. 1939, G.B. Fairchild. (11) fUSNM, Rio Anton in CURACAO: No specific locality (Holthuis, El Valle Crater, 66* (11.2-21.0), 1$ (13.9), 7 Apr 1980:69). (1) fRNHL, stream in mango planta- 1962, HL. (12) fUSNM, Rio Arenal at Carretera tion at Santa Cruz, 226* (6.1-21.7), 11$ (5.7-12.9), Interamericana, 56* (9.1-11.3), 1$ (15.4), 5 ovig 2 ovig 9 (11.0-11.7), 12juv (3.1-5.9), 4 Jan 1957, $ (10.1-16.2), 13 Oct 1961, HL. (13) fUSNM, L.B. Holthuis; RNHL, 646* (7.5-23.6), 30$ Rio Guabas, 1.7 km W of Anton, 76* (8.6-21.9), (7.2-13.6), 14 ovig $ (10.9-16.0), 67 juv (3.0-8.0), 3$ (10.3-16.6), 6 ovig $ (13.3-15.6), 14 Oct 1961, 11 Feb 1957, LBH; RNHL, 416* (8.1-20.9), 22$ HL. (14) fUSNM, Rio Arenal 5.1 km E of San (5.8-14.5), 3 ovig $ (11.3-13.4), 8 May 1957, Carlos, 1$ (16.2), 11 Mar 1962, HL. (15) fUSNM, A.CJ. Burgers, F.L. Hermans. (2) fRNHL, Rio Las Lajas E of San Carlos, 16* (19.0), 5$ stream near Hato airport, 206* (12.1-23.4), 18$ (8.8-16.4), 24 Mar 1962, HL. (16) fLGA, El (9.2-21.8), 36 ovig $ (10.2-14.9), 10 Jul 1957, Valle, 26* (10.5, 12.1), 2 ovig $ (14.2, 21.1), 5 juv ACJB. (4.3-5.5), May 1978, LGA, B.E. Felgenhauer. NICARAGUA: (1) fPM, Hacienda El Polvon, De- (17) fLGA, El Valle, Mosa Cham, 2$ (14.5, 15.3), partamento Occidental (Smith, 1871:95), syn- 14 Apr 1973, LGA; LGA, 96* (7.1-11.4), 14$ types of Atya tenella, 16* (16.2), 1$ (9.4), summer (5.8-17.9), 14 Apr 1973, LGA. (18) fLGA, Rio 1868, J.A. McNiel; MCZ, 4$ (11.2, 13.7, 13.7, Anton, El Valle, 1 ovig $ (14.5), 30 Jan 1974, 18.2), 1871(?), J.A. McNiel, (2) fMCZ, Corcuera, D.C. Darling. Provincia de Colon— (19) fLGA, 72 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Cano Rey, 12 juv (about 6.0), 18 Aug 1972, LGA. of the acumen is marked by angles (with rounded Provincia de Darien—(20) fUSNM, trib to Rio apices) of 75 to almost 90 degrees. One to three Chucunaque, Yaviza, 46* (13.5-21.5), 3$ small preapical spines may or may not be present (11.1-18.1), 16 Feb 1924, J.L. Baer. Golfo de Chi- on the ventral keel. There is some variation in the riqui—(21) fLGA, Isla Secas, 9juv (less than 5.0), height of the dorsal keel, but it is never so concave Feb 1974, P. Glynn. Provincia de Panama—{22) along the median part of its length as to dip fUSNM, Cerro Azul area, 306* (9.6-26.4), 179 below the level of the lateral carinae. The length (7.1-19.4), 2 ovig 9 (19.0, 19.4), 29 Jan 1971, of the acumen is also variable, and although there LGA. (23) fLGA, first stream on Carti Rd, 11 is no consistent correlation between its length and km, 26 (13.3, 13.9), 15 Jan 1975, Kramer. Archi- that of the carapace, generally the smaller indi- pielago de las Perlas— (24) *LGA, Isla Pedro Gon- viduals have relatively longer ones, and the short- zalez (Abeleand Blum, 1977:240), 2$ (13.3, 13.4), est are found in the larger specimens. 13 Jun 1973, LGA; LGA, 3d4 (14.2, 15.9, 17.7), 26 The pterygostomian angle, although always Mar 1976, Kramer. (25) Isla San Jose (Abele and acute, may or may not be somewhat produced. Blum, 1977:240), 26 (5.7, 9.5), 1$ (4.9), 4 juv, 20 The pile of setae on the carapace is best developed May 1973, LGA. Provincia de San Bias— (26) in animals that have recently molted. Even in fMCZ, San Bias, \6 (15.6), 6 Jun 1929. (27) them some magnification is essential to render fLGA, 20.4 km N of El Llano on Carti Rd, 26 them visible, and the exoskeleton is so smooth (7.2, 12.3), 10 Mar 1973, R.L. Dressier, McPhail. that one has difficulty in holding a living speci- Provincia de Veraguas—(28) fUSNM, creek 8.5 km men. The difficulty no doubt increases with time W of Sona on Remedios Rd, 36 (10.3-17.5), 29 following a molt, for in many of the shrimp much Oct 1961, HL. (29) fUSNM, Rio Martin Grande of the carapace and abdomen appear to be devoid 6.8 km S of Santiago, 3$ (11.8-16.0), 14 Jun 1962, of setae. HL. (30) fUSNM, trib to Rio Tabasara 3.4 km The height of the body, particularly in the area W of El Maria Creek on Sona-Remedios Rd, 16* (25.0), 3$ (9.4-16.0), 1 ovig? (15.5), 29 Oct 1961, HL. (31) fUSNM, Rio Veraguas, 11.9 km W of El Maria, 16* (22.6), 11 Nov 1961, HL. (32) fLGA, Rio Las Guias at Rd from Calobre, 16* (about 6.0), 21 Feb 1973, LGA, MHR. Province unknown— (33) fLGA, Cerro Jefe, 16* (22.8), 1969, RLD. (34) fLGA, Cerro Jefe at 635 m, 2 ovig 9 (18.0, 19.2), 29 Apr 1973, LGA, RLD. ERRONEOUS RECORDS.—(1) New Caledonia (A. Milne-Edwards, 1864:148) for A. robusta. (2) Oceania (Oliveira, 1945:179) for A occidentalis. VARIATIONS.—Among the most conspicuous variations in Atya innocous is the form assumed by the rostrum (Figure 29). In individuals with car- apace lengths of five to seven mm, the margins are strongly tapering, sometimes with hardly a trace of the anlagen of the paired subangular C d bends that mark the base of the acumen in larger individuals and that become more prominent FIGURE 29.—Atya innocous, variations in rostrum and eyes with increase in size of animal (scaled to approx. same size; following subsequent molts. In most individuals numbers in parentheses = carapace length in mm): a, male with carapace lengths of 20 mm or more, the base (7.8); b, male (9.9); c, female (15.1); d, male (24.7). NUMBER 364 73 of the first and second abdominal segments, is ure 23/), but in other individuals from the same quite variable, and in most of the material from locality the denticles may be aligned in two series, the Pacific watershed it is proportionately higher or they may be more irregularly dispersed (Figure than it is in specimens from the streams draining 23*). into the Gulf of Mexico and Caribbean, but Many other variations have been noted, but among specimens from the latter drainage basins none of them has enabled us to distinguish all there are those that cannot be distinguished from members of any population from at least some forms frequenting the Pacific versant. The ab- individuals of populations occurring elsewhere. A sence of denticles on the ventral margin of the comparison of the illustrations presented herein third through fifth abdominal pleura will often will provide an appreciation of some of the dif- serve to distinguish specimens from the latter ferences and the many similarities existing in basin, but frequently individuals from streams populations from several parts of the range of this flowing into the Gulf or Caribbean lack such shrimp. denticles. In a series of specimens from Mannet's As pointed out above, populations of this Gutter, a tributary of the Layou River on Dom- shrimp occurring in streams emptying into the inica, denticles are present on the ventral margin Pacific usually have bodies that are, or seem to of the pleura of the third through, fifthabdomina l be, more highly vaulted than those in the Carib- segments in most. In others, they are asymmetri- bean-Gulf Basin. They also possess more spindly cally present on the fourth only, present on the legs than we thought at first to be typical of the fourth and fifth, a single one on each of the fifth, more eastern members of the species. Later we and occasionally absent from all of them. found that these were the only two traits on which The proportions of the distal four podomeres we were relying to distinguish A. tenella (Pacific) of the third pereiopod are highly variable, and from A. innocous (Caribbean-Gulf-western Atlan- except for the observation that generally the pro- tic), and as more specimens were compared we podus in larger males is broader than in most discovered spindly legged individuals with highly other individuals, no generalizations relative to vaulted carapaces among those insular popula- their structure seem possible to us. Most differ- tions in the Antilles, thus indistinguishable from ences appear to be individual ones that oftentimes specimens from the Pacific versant. To be sure, are not symmetrical. We have given special con- we have encountered none from the latter area sideration to the ornamentation--. of* the flexor that have as robust pereiopods as exist in many, surface of the propodus and dactyl,! and some of if not most Antillean representatives of the spe- the variations noted are depicted in Figures 22i, cies, but to recognize A. tenella as a distinct taxon 23^, 30. To be sure, the development of the setae under these circumstances seems to us to be in- flanking (sometimes almost surrounding) the defensible. spines are best developed in individuals that have ECOLOGICAL NOTES.—The most complete ac- recently molted, and in some specimens that were counts of the habitat and habits of this shrimp preserved in late intermolt stages the setae are are those of Chace and Hobbs (1969) and Fryer indeed few, presumably lost through abrasion. (1937)\ The former began their discussion of its The degree of sclerotization of the spines in- ecological distribution (p. 60) with the statement creases, within limits, following .a molt, but the that size and shape of the spines seem, not:; to be correlated with either growth or stages of the molt Atya innocous is probably the least ecologically and geograph- cycle, and certainly not with different parts of the ically restricted shrimp on the Island of Dominica.... It seems to be equally at home in the cascading reaches of range of the species. rivulets [at altitudes of approximately 800 me- The dactyl of the fifth pereiopod in some pop- ters], in quiet upland pools, and in low-lying sluggish brooks. ulations is provided with a pectinate comb (Fig- In an upland pool ... some 20 or 30 individuals were 74 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 30.—/40u innocous, variations in flexor surface of distal part of third pereiopod (numbers in parentheses = carapace length in mm): a, Saint Croix, Virgin Islands (<5, 20.4); b, Banana River, Jamaica (6, 18.3); c, Banana River, Jamaica (6, 22.2); d, Rio de los Platanales, Costa Rica (5, 17.5); e, El Valle crater, Panama (6, 20.7);/ Isla Barro Colorado, Panama (9, 19.4); g, 3.4 km W of El Maria, Panama (6, 25.0). NUMBER 364 75 observed actively crawling over the bottom at about 10:00 and reappearing shortly thereafter. Atya innocous was by far A.M. the most active. Of the three or four large individuals of M. crenulatum, two were in view but were not moving about, and The shrimp ventured into full sunlight but the others were concealed somewhere among the stones. quickly returned to the shaded area of the pool. Except for the almost incessant wanderings of A. innocous and Fryer (1977:69) wrote the smaller M. crenulatum, there was little activity. When an earthworm, suspended on a string, was gently that its tolerance exceeds the range of natural conditions on lowered into the water, however, a chain reaction was initi- Dominican streams was indicated by its occurrence in a ated that set the entire population of the pool in motion. pond whose bottom was composed largely of soft ooze. When The smaller individuals of M. crenulatum were the first to this pond, which had been constructed on a stream, was show an awareness of the presence of the worm. Although drained, hundreds of individuals were collected for human apparently they did not see it, the rate of their random consumption. walking increased tremendously, perhaps best described as The following quotation from Chace and Hobbs "frantic," causing them to collide with one another and with the other inhabitants; A. innocous joined them, the larger M. (1969:43-45) provides some insight into its niche crenulatum began moving back and forth, and X. elongate left in a small community occupying a pool on Man- the stones on which they had been comparatively still and net's Gutter, Clarke Hall Estate, Dominica. swam about the pool. The frenzied motion disturbed the fish and they, too, began to shift their positions on the rocks. A number of hours were spent in observing the decapod When one of the smaller M. crenulatum finallylocate d the fauna of a small pool, approximately 4 feet wide, 10 feet position of the worm, the shrimp left the bottom, swam to long, and with a maximum depth of 2 feet.... Here, at an the worm, grasped it with its chelae, and attempted to swim altitude of some 350 feet, the stream has cut a deep V-shaped away with the worm. When this attempt failed, the shrimp valley, and boulders from the eroded walls have fallen into used its abdomen to give a rapid series of strong tugs. This the stream bed, deflecting the current or obstructing the motion apparently attracted the attention of almost all of flow so that pools are interspersed between cascades and the shrimps in the pool. One of the larger M. crenulatum swam riffle areas. Except after heavy rains the water is clear, toward the worm as the smaller ones backed away, and flowing over a rocky, sandy bottom overlain in the pools when it had stripped the worm from the string, the shrimp with a shallow layer of silt. Many of the larger trees along sank to the bottom of the pool and scurried for the nearest the 45 to 80 degree slopes have been cut and bananas have cover. By this time, both individuals of M. carcinus had been planted between the felled trunks; as a result, during moved into the open water of the pool, and as the larger one a short period in the day, the time depending upon the moved about, all of the other shrimps retreated from its orientation of the adjacent slopes, full sunlight reaches the path, remaining beyond reach of the large chelipeds. While stream bed. Marginal shrubs do inhibit direct light from the commotion was going on, the crab slowly crawled from reaching portions of the stream. The pool referred to here under the largest rock at the side of the pool, and all of the received sunlight for about two hours before noon in mid- occupants, including the previously dominant M. carcinus March. gave wide clearance to the newcomer. A short time after the shrimp with the worm had found cover and presumably had The macroscopic fauna of the pool consisted of what the devoured it, the crab crawled back into its lair, the large M. observer believed to be a single species of gobioid fish scarcely carcinus moved into crevices, and the remainder of the pop- exceeding 40 mm in length; two atyid shrimps, Atya innocous ulation returned to its original state. A second worm intro- and Xiphocans elongate; two palaemonid shrimps, Macrobrach- duced into the pool resulted in a similar turmoil, but this ium carcinus and M. crenulatum; and a crab, Guinotia dentate. worm was successfully acquired by one of the young M. No other animals were observed. There were 1 crab, 2 M. crenulatum, which quickly swam to the shallow down-stream carcinus, 12-15 A. innocous, about the same number of M. end of the pool and crawled beneath a stone. crenulatum, 20-25 Xiphocans elongate, and approximately the same number of fish. When the pool was first approached by observers, the fish According to Fryer (1977:69), Atya innocous is and X. elongate were "resting" on the stones at the sides and basically an "ambulatory species," rarely swim- bottom of the pool, only swimming occasionally for a centi- ming unless disturbed, and its ability to exploit meter or so to take another position on a rock. The crab was the wide range of habitats mentioned above "can not in sight, and only the chelae of the smaller of the two M. be attributed to a considerable extent to its ability carcinus could be seen protruding from beneath one of the larger rocks. In contrast, the individuals of A. innocous and to collect food by two very different methods": the smaller M. crenulatum were scurrying back and forth (1) filtering and (2) sweeping and scraping, utiliz- across the bottom of the pool, disappearing beneath a stone ing the first two pairs of pereiopods. It also tol- 76 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY erates a wide range of temperature. It has been containers up to six days. The second larval stage, found in water at 21°C, and "in aquaria it has age three to eight days, was 1.85 to 2.05 mm in remained healthy in spite of falls in temperature length. The third larval stage, age seven to 13 of 14°C . .. , individuals from running water days, attained lengths of 2.0 to 2.15 mm. The survived for over two weeks in confined conditions fourth larval stage, age 12 to 17 days, exhibited in which temperatures rose to at least 32°C, lengths of 2.4 to 2.7 mm, and the endopod of the withstood a trans-Atlantic air journey, and then first pereiopod had become 5-segmented. Follow- flourished in aquaria." ing the fourth stage, "morphological changes ac- Populations in pools were found by Fryer to be companying moults were less marked," but de- active both at night and during the day, but their scriptions and illustrations of 12 stages are pro- nocturnal activity is greater. He observed: "By vided, the last of which was reached at an age of night many feeding animals frequented the walls 76 to 119 days at lengths of 10.05 to 11.75 mm. of the pool which were almost deserted by day" Metamorphosis into juveniles followed. (p. 70). Details of grooming, ambulation, feeding, Fryer (1977:72) reported: "In aquaria adults and the functional morphology of the foregut are readily mated and produced fertile eggs from also described by Fryer (pp. 94-97,111-118). The which larvae emerged, but never survived. Even food appears to consist largely of "paniculate experienced prawn-rearers at Conway failed to detritus" derived from vegetation overhanging or keep these for more than 10 days." Fryer sug- adjacent to the streams, but filamentous algae gested that unsuitable food or adequate space firmly attached to rocks were also successfully might have been responsible for their demise. In removed and presumed to have been eaten (p. view of Hunte's success in rearing larvae, perhaps 96). the failure was due to the necessity for them to At night, populations in a stream on £1 undergo their larval development in saline water. Yunque, Luquillo National Forest, Puerto Rico One of the specimens taken to England by (altitude 610 m) were observed by one of us Fryer (1977:72) "lived for about 6 years and 9 (CWH) to arrange themselves in rows facing months in an aquarium and had not achieved its upstream in shallow riffle areas. It is assumed maximum size at the time of death. Several others that this behavior facilitates the capture of detri- lived for 3 or 4 years." tus and planktonic organisms. Chemical and Abele and Blum (1977:245) reported two ovi- physical data recorded for this stream in June gerous females from the Perlas Archipelago with 1962 (Hart, 1964:332) included a pH of 7.2; total total lengths of 45.1 and 45.3 mm. They were hardness, 34.2 ppm; SO4,0.4 ppm; Cl, 10.0 ppm; carrying 2100 and 3650 eggs (0.6 to 0.7 mm in and a temperature of 24.8° C. length). LIFE HISTORY Notes.—Ovigerous females hav- COMMON NAMES.—We are aware of no vernac- ing carapace lengths of 7.9 to 24.9 mm, collected ular names that apply only to this shrimp, but at during every month except September and No- least some of those listed for Atya scabra are not vember, are present among the specimens exam- specific and are used to designate either or both ined by us. Bonnelly de Calventi (1974b) found species. See the common names of A. scabra below. an ovigerous female in Canada Madrigal, Do- The FAO names cited by Holthuis (1980:69) minican Republic, in February. Hunte (1979b), for this species are Basket shrimp (English), Sal- basing his study on larvae reared in the labora- tarelle panier (French), Camaron cestillo (Span- tory, found that they thrive best in a salinity of ish). 30%o. The eggs hatched, most frequently at night, into larvae 1.8 to 1.9 mm in length bearing Atya intermedia Bouvier the typical series of appendages from FIGURES 1/, 9, 10, 12*/, 31, 32 the antennules through the third maxillipeds. Atya (sp. nov.?).—Bouvier, 1889:82. These larvae did not feed and were present in his Atya intermedia Bouvier, 1904:137, 138 [type-locality: "Af- NUMBER 364 77

rique occidentale: Tie Saint-Thomas"; types: not extant]; presented illustrations of the rostrum and anten- 1905:110, 112, 119, 120, fig. 23; 1906:493; 1925:292, SOS- nal spine, and pointed out the same features SI 3, 322, 323, 356, figs. 690-695.—Holthuis, 1951:24; 1966:235-237, fig. 5a-d.—Monod, 1967:110, 119, 135, pi. included in his original description; he also in- ix: figs. 18-24.—Lemasson, 1973:68. cluded measurements of the legs, comparing them with those of members of A. robusta (= Atya inno- REVIEW OF LITERATURE.—Bouvier (1889), in cous), A. scabra, and A. africana. his study of the nervous system of decapod crus- In 1906, Bouvier reported that among a small taceans, stated that collection from the same island were three large j'ai eu l'occasion d'observer un certain nombre d'Atya (sp. and several small specimens of A. intermedia, the nov.?) grace a l'obligeance d'un jeune savant portugais, M. smallest of which are about the size of the "Car- Nobre, qui avait recu du Maroc un certain nombre de ces idenes" and have almost triangular rostra. The animaux. Ce Crustace est une grande crevette africaine, et account of the species in Bouvier's monograph il ne differe des Palemons asistiques etudies par Milne- (1925) is somewhat expanded, and more infor- Edwards que par des caracteres tout a fait secondaires. mation is given concerning the types, which were As Professor Holthuis has pointed out to us, males ranging in carapace length from 62 to 82 evidently Bouvier made two errors: (1) in stating mm. Especially emphasized were the variations that the animals had been obtained in Morocco, in the presence or absence of prominent articu- and (2) in using the term "Palemons" instead of lated spines and of the ornamentation of the Atyoidees to designate the shrimp mentioned by dactyl on the ambulatory legs. He also added a A. Milne-Edwards (1864:145-152). Holthuis (in more precise locality, the Ouro River, for the letter) noted further that inasmuch as (1) the only species on Sao Tome and presented several new Atya received by the Paris Museum from Nobre illustrations. Holthuis (1951) added no new infor- are the type specimens of A. intermedia (see Bou- mation but in 1966 described specimens from an vier, 1904:137, 138), (2) these specimens were effluent of Crater Lake on Annobon, comparing large, (3) A. Milne-Edwards did not publish in- them with A. sulcatipes. He referred to Bouvier's formation on Asiatic palaemonids but did on comparison of A. intermedia with A. africana, agree- Asiatic atyids, and (4) neither large atyids nor ing that the two species are distinct, and further large palaemonids occur in Morocco, there is noted the relationship to A. innocous. Illustrations good reason to believe that the "Atya (sp. nov.?)" of the cephalic region and observations of the were the types of Atya intermedia. (Unfortunately, color of the Annobon specimens were also in- these specimens could not be located when one of cluded. Monod (1967) added no new data. Le- us (CWH) visited the Museum National masson (1973), considering native African d'Histoire Naturelle, Paris.) The original descrip- shrimps suitable for cultivation, referred to this tion of this shrimp was prepared from four spec- insular species and stated that it attains a length imens, and Bouvier characterized the species as of eight centimeters. possessing a dorsal carina that reaches the tip of PUBLISHED ILLUSTRATIONS.—The earliest illus- the rostrum but with a complete atrophy of the trations of the species are those of Bouvier (1905) ventral rostral carina, and by the rounded prom- consisting of a dorsal view of the rostrum and a inences (very weakly squamiform and sparingly lateral view of the rostrum and antennal spine. corneous) on the meropodites of the (presumably The most complete series is that included in his third through fifth) pereiopods. He stated further monograph (1925): a dorsal view of the rostrum, that the rostrum is large and little lengthened a lateral view of the cephalic region, a dorsal view anterior to the lateral notches, which are strongly of the telson, either cephalic or caudal views of obtuse, and that its legs are much slenderer than the first and second pleopods of the male, and those of any other large forms of the genus. illustrations of the dactyls of the third pereiopods Bouvier (1925) included this shrimp in his key to of two specimens. The figures presented by the then known species assigned to the genus, Holthuis (1966) are dorsal and lateral views of 78 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY the cephalic region of an adult and of a juvenile. nule; dorsal median carina, lying dorsal to lateral Monod's (1967) illustrations were taken from carinae except anteriorly, rather gently curved Bouvier (1925). (not concave) almost to apex of rostrum where DIAGNOSIS.—Cephalic region of carapace not suddenly bent ventrally, disappearing just before conspicuously sculptured; spines limited to anten- reaching apex of acumen; ventral carina with 2 nal, pterygostomian, and occasional small ventral subapical minute teeth; ocellar beak with anterior rostral. Rostrum with margins rather suddenly margin directed anterodorsally, forming acute contracted forming angle anterior to orbit, angle angle with dorsal margin and reaching level of never produced; dorsal surface without median distal margin of eyes, thus falling far short of tip row of strong spines. Ventral margin of second of stylocerite; dorsal margin abutting longitudi- through fifth abdominal pleura without rows of nal ventral groove posterior to ventral carina of corneous spinules, and caudoventral angle of rostrum. Antennal spine well developed; ptery- fourth and fifth pleura acute to subacute but not gostomian angle rather weakly produced, not produced in spines. Sternum of fifth abdominal spiniform; no spines present between antennal segment with small median tubercle, sternum of and pterygostomian spines. Surface of carapace sixth about 1.2 times as broad as long. Preanal with many small setiferous punctations, but de- carina with caudoventrally directed spine and void of ridges and spines other than those just smaller spine (or at least bulge) between spine mentioned; all setae on carapace small and in- and base of carina, major spine overreaching conspicuous. basal part of sclerite bearing it. Telson about 1.6 Pleura of first 2 abdominal segments (Figure times as long as wide with approximately 6 spines 31/) rounded posteroventrally; third subangular, in each of 2 dorsal rows. Antennular peduncle and fourth and fifth distinctly angular, latter 2 with dorsal surface of proximal article devoid of although acute not produced. All pleura lacking sclerotized spinules proximal to transverse distal corneous spinules and conspicuous fringe of setae row; penultimate article 1.3 to 1.5 times as long on ventral margins. Fourth abdominal tergum as wide, and dorsal surface with many small about 1.2 times as long as fifth which subequal in scattered corneous spinules. Coxae of third and length to sixth and to telson. Sternum of fifth fourth pereiopods lacking anterolateral spines. abdominal segment bearing very small com- Third pereiopod with merus rounded ventrally, pressed tubercle at midlength (Figure \f). Ster- 3.7 to 4.0 times as long as high, ventromesial num of sixth segment about 0.8 as long as wide. surface bowed, never parallel to that of corre- Free part of preanal carina (Figure 31*) projecting sponding podomere of other member of pair, and far beyond basal part of sclerite, underlying lateral surface bearing corneous tubercles, most smaller accessory spine. Telson (Figure 31/) about of those on merus and carpus with scalelike ex- 1.6 times as long as wide, its dorsal surface bearing tremities, and most arranged in linear series; pro- paired concave rows of 6 corneous denticles; pos- podus almost 3 times as long as wide, flexor teromedian tubercle not overhanging caudal surface with corneous spiniform tubercles distrib- margin of telson. uted in sublinear series, spines not contiguous; Proximal podomere of antennule (Figure 316) dactyl not fused with propodus and bearing 2 with stylocerite reaching base of distal fifth of linear or sublinear series of denticles on flexor segment; dorsal surface with linear cluster of setae surface. but lacking corneous spinules; distal margin bear- MALE (Annobon Island).—Rostrum (Figure ing row of 11 corneous spinules; penultimate 3la,d) with margins weakly concave and slightly segment of peduncle about 1.4 times as long as tapering to subangular bend at base of acumen, wide and bearing 13 (right) or 14 (left) spinules latter almost reaching articulation between on dorsal surface and 16 on distal margin; ulti- proximal and penultimate podomeres of anten- mate podomere with row of 11 spinules at base of NUMBER 364 79

FIGURE 31.—Atya inttmudia (all from male from Annobon): a, dorsal view of cephalic region; b, dorsal view of antennular peduncle; c, mesial view of appendices masculina and intema; d, lateral view of cephalic region; e, lateral view of preanal carina;/, lateral view of second through fifth abdominal pleura; gjt, flexor surface of distal part of fourth and fifth pereiopods, respectively; i, lateral view of third pereipod; j, dorsal view of telson; k, flexor surface of distal part of third pereiopod. (Scales marked in 1 mm increments.) 80 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY lateral flagellum, 4 at base of mesial one, and 2 distal third of carpus of third pereiopod; length proximal to lateral row. Antenna with ventrolat- of merus 1.9 times length of carpus, and latter eral spine on basis falling a little short of tip of almost as long as propodus. Fifth pereiopod stylocerite; lateral spine on scaphocerite strong, reaching end of proximal fifth of propodus of reaching end of proximal third of ultimate pod- fourth pereiopod; merus 1.4 times as long as omere of antennular peduncle; lamella far over- carpus, latter about 0.7 as long as propodus. reaching latter; flagellum extending to end of Ornamentation of merus, carpus, and propo- anterior fourth of telson. dus of fourth pereiopod similar to that of third Third maxilliped overreaching antennular pe- except for addition of ventrolateral spine on duncle by about half length of distal podomere merus at base of distal fourth and ventral one at of endopod; tip of exopod reaching base of distal midlength, and 1 ventrolateral and 2 distolateral sixth of penultimate podomere of endopod; latter ones on carpus. Ornamentation of fifth pereiopod only slightly (less than 1.1) times as long as like that of fourth but with 2 ventral spines on ultimate podomere. merus proximal to ventrolateral spine, and 3 First pereiopod reaching level of end of proxi- distolateral spines on carpus. mal third of ultimate podomere of antennular Diaresis of lateral ramus of uropod flanked peduncle, second reaching base of distal third of proximally by row of 21 (right) or 22 (left) artic- fingers of first pereiopod; terminal brush of setae ulated, corneous denticles and slightly larger fixed of both appendages lacking scraping denticles. spine at lateral end of row. Third pereiopod (Figure 3lijc) without spines on COLOR NOTES.—The only observations on color merus and carpus, and when extended anteriorly, are those of Holthuis (1966:237): "The whole overreaching antennular peduncle by dactyl and body is of a more even dark greenish grey color; four-fifths of propodus; merus with ventromesial it is not marbled as in A. sulcatipes. There is a light margin bowed, about 3.7 times as long as high median spot on the fourth and sixth abdominal and 2.4 times as long as carpus; propodus 3.5 somites. The latter does not show the dark trans- times as long as wide, subequal in length to verse band so conspicuous in A. sulcatipes." carpus; distoventral margin of coxa entire and SIZE.—The largest specimen reported was that mesial caudoventral prominence rudimentary recorded by Holthuis (1966:235); it had a total and lacking conspicuous setal clusters. Lateral, length of 83 mm, and presumably its carapace dorsal, and ventral surfaces of merus studded length was the maximum he recorded, 32.0 mm. with many small sclerotized (at least apically) The largest female measured by us had a cara- tubercles, most arranged in linear series; clusters pace length of 19.1 mm. The smallest and largest of plumose setae flanking tubercles, most conspic- ovigerous females had corresponding lengths of uous tufts forming oblique row on lateral surface; 7.9 and 15.4 mm. mesial extremity of podomere very weakly pro- DISTRIBUTION AND SPECIMENS EXAMINED.—The duced at level of mesial articular condyle of car- range of Atya intermedia is not known to extend pus. Carpus strongly tuberculate except on flexor beyond two islands in Gulf of Guinea: Annobon surface where tubercles absent; tufts of long plu- (= Pagalu) and Sao Tome (Figure 32). mose setae present on flexor surface of podomere. Records for the known localities are listed be- Propodus with tubercles on extensor, mesial, and low. Collections that we have examined are lateral surfaces arranged mostly in linear series, marked with an asterisk. Numbers following the rows less well defined on flexor surface. Dactyl specimens listed are measurements, in mm, of the movable, its flexor surface with 2 rows of denti- carapace length or, if followed by "t.l.," total cles, proximal one of 4 more mesially situated length. Some listings lack dates and/or collectors; than lateral one of 3. if so, these could not be determined. Fourth pereiopod with dactyl reaching base of ANNOBON: *RMNH, Annobon (Holthuis, 1966: NUMBER 364 81

FIGURE 32.—Distribution of Atya intermedia (circled numerals = number of localities).

235) 16* (14.3) 1959, R. Zariquiey Alvarez. *BM, collected from a river (Bouvier, 1925:309) and Annobon, 46 (11.2-16.0), 15$ andjuv (7.3-19.1), others from an "effluent near western shore of Jul-Aug 1959, Cambridge University President's [Crater] lake" (Holthuis, 1966:235). Expedition. (1)*RMNH, Crater Lake, Annobon, LIFE HISTORY NOTES.—The only ovigerous fe- Pillsbury sta 278 (Holthuis, 1966:235), 86 (8.5- males known are the four that were collected on 26.7), 9$ (8.0-9.4), 4 ovig 9 (7.9-15.4), 87 juv 20 May 1969. No other aspects of the life history (3.0-8.7); *USNM, 36* (8.5-25.4), 1 juv (5.6), 20 of the members of this species have been recorded. May 1969. REMARKS.—Atya intermedia is very closely allied SAO TOME: "Saint Thomas" (Bouvier, 1904: to A. innocous and is more remotely related to A. 138), 46 (62-82, t.l), 1895, M. Nobre. (1) lanipes. The similarities are best demonstrated in *MHNP Rio do Ouro (Bouvier, 1925:309), 26 the rostrum which has margins that are angular (20.5, 24.5), 2$ (8.0, 10.2) 1 sex? (19.5), M. Gra- to subangular at the base of the acumen; the vier. basal podomere of the antennule lacks sclerotized ECOLOGICAL NOTES.—The only data available denticles on the dorsal surface proximal to the are the statement that some of the specimens were distal transverse row; the lateral and dorsal sur- 82 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY faces of the merus, carpus, and propodus of the 20-22, 26-36, 38-40, 47, 49, 52-54, 59, figs.6-9.—Hobbs , third pereiopod bear conspicuous corneous tuber- Hobbs, and Daniel, 1977:150.—Fryer, 1977:124. Atya Lanipes.—Bonnelly de Calventi, 1974a: 16. cles, and the tubercles on the flexor surface of the dactyl are arranged in two rows; the ventral REVIEW OF LITERATURE.—Holthuis (1963) margin of the abdominal pleura are devoid of named this species on the basis of two male corneous spinules, and the sternum of the fifth specimens that had been obtained about 1898 by abdominal segment and the telson exhibit ap- the Rijksmuseum van Natuurlijke Historie from proximately the same proportions as those of A. Saint Thomas, Virgin Islands. Five years follow- innocous. Indeed the difference between the pop- ing the appearance of Holthuis' description, ulations of Atya on Annobon and Sao Tome and Chace and Hobbs (1969) included the species in the American A. innocous is of such a minor nature their key to freshwater and terrestrial decapods that had these insular African populations not of the West Indies, recorded its presence in eight already received a name, we should treat them as localities on Puerto Rico, discussed its range and a regionally restricted variant of the latter species. variations, and illustrated the distal part of the Although Bouvier (1925:309, 310) compared second pleopod of the male. Chace (1972) added his specimens of A. intermedia with those of A. no new data for the species. Bonnelly de Calventi africana and the type of A. robusta, comparing them et al. (1973) recorded this shrimp from the Rio with specimens he identified as A. occidentalis (» Canada Madrigal, Dominican Republic, and A. innocous) would have been more rewarding. Bonnelly de Calventi (1974a) reported the cara- This is especially obvious in examining the ratios pace lengths of her specimens (16*, 18.9 mm; 7$, cited (p. 311), among which most of those of A. 11.7-17.2 mm); the locality listed is recorded here intermedia and A. occidentalis overlap. under "Distribution and Specimens Examined." The only really distinctive feature of this This work was expanded (1974b), including a shrimp that we have encountered is the occur- diagnosis of the shrimp, illustrations, notes on rence of an accessory spine, or small but clearly color, size, habitat, and distribution in the Do- defined, rounded prominence on the preanal car- minican Republic, where it is not very common ina situated just dorsal to the larger caudoventral in the southern part of the country. Peck (1975) spine. listed Atya lanipes among the crustaceans frequent- The rostrum of small juveniles, as in all of those ing caves on Jamaica. In describing Atya dressleri, species in which the adult exhibits an angular Abele (1975) stated that A. lanipes is its closest rostrum, is more nearly acuminate and seems to relative and cited several characters in which they become less so with each molt until in the adults differ. Considerable additional information con- the angles at the base of the acumen approach or cerning this shrimp was provided by Hunte surpass 80°. (1975) in his study of members of a population occurring in the Cane River, Saint Thomas Par- ish, Jamaica. He presented color notes, discussed Atya lanipes Holthuis morphological peculiarities and variations in the FIGURES la, 4b,c, 9, 10, 12?, 33-36 Jamaican specimens, and described and illus- trated the first larval stage. Atya lanipes Holthuis, 1963:61-67, figs. 1, 2 [type-locality: Saint Thomas, Virgin Islands: types: RMNH, 6* holotype, Villamil and Clements (1976) added many 6 paratype]; 1977:271, 272, 1 fig.—Chacean d Hobbs, interesting data on the shrimp in the upper Rio 1969:5, 14, 19, 29, 30, 33, 57, 61-63, 73, fig.14c—Chace , Espiritu Santo, Puerto Rico, including observa- 1972:14.—Bonnelly de Calventi et al., 1973:1338.—Bon- nelly de Calventi, 1974b:35, 38, 39, figs. 5, 12.—Peck, tions on habitat, feeding, size, and life history. 1975:308.—Abele, 1975:56.—Hunte, 1975:66-72, figs.1 - Holthuis (1977) reported the presence of this 3; 1978:135, 136, 139, 144, 145, 147, fig. 7; 1979a: 153; shrimp in two localities in Cuba. He contrasted 1979b:240; 1979c;70.—Villamil and Clements, 1976:3, 4, an ovigerous female with the types and figured NUMBER 364 83 the distal podomeres of the fifth pereiopod, point- spines other than antennal and pterygostomian; ing out the spine on the distal flexor extremity of ventral spines on rostrum usually present, ptery- the propodus that opposes the dactyl. Hobbs, gostomian spine strong; rostrum with margins Hobbs, and Daniel (1977) only reiterated the tapering from base to apex, never bearing lateral occurrence of this shrimp in spelean waters on angles. Ventral margin of all abdominal pleura Jamaica noted by Peck. Among the more recent without rows of sclerotized denticles and caudo- references to the species is that of Fryer (1977) ventral angles of fourth and fifth pleura not who pointed out that "Atya lanipes and Potimirim produced in spines. Sternum of fifth abdominal americana (Guerin-Meneville) are also known only segment with small, rather inconspicuous, me- from the West Indies, each being reported from dian tubercle, that of sixth slightly longer than three islands of which Jamaica is common to broad. Preanal carina without or with very short both." A discussion of the geographic and eco- spine. Telson 2.1 to 2.6 times as long as wide and logical distribution of A. lanipes on Jamaica was with 5 to 7 spines in each of 2 dorsal rows. presented by Hunte (1978), and he (1979a) re- Antennular peduncle with proximal article lack- ferred to his earlier description of the first larval ing premarginal sclerotized spinules dorsally; pe- stage. In a second paper that year (1979b), he nultimate article about twice as long as wide and pointed out the error in his interpretation of bearing very few, if any, small spinules, if present, rudiments of the first maxillipeds. He concluded sometimes arranged in linear series. Coxae of that, like Atya innocous and Micratya poeyi (Guerin- third and fourth pereiopods lacking prominent Meneville), the first larval stage of A. lanipes lacks ventrolateral spine; that of third without mesial periopods. This species was included by Hunte caudoventral prominence. Third pereiopod with (1979c) in his list of the atyid and palaemonid merus rounded ventrally, 6 to 8 times as long as shrimps of Jamaica. high, ventromesial surface bowed, never parallel PUBLISHED ILLUSTRATIONS.—The illustrations to that of corresponding podomere of other third presented by Holthuis (1963) include dorsal and pereiopod, and lateral surface lacking sclerotized lateral views of the cephalic region, postaxial view tubercles and spines; propodus 3 to 4 times as of the third maxilliped, caudal view of the first long as broad, extensor surface rugose but lacking pleopod of the male, caudomesial view of the heavily sclerotized tubercles and spines, flexor second pleopod of the male, dorsal view of the surface with small spines often linearly arranged, telson, and illustrations of the first, third, fourth, spines partly surrounded (sides and distally) by and fifth pereiopods. Chace and Hobbs (1969) usually inconspicuous short plumose setae; dactyl presented a posteromesial view of the second freely movable and bearing 2 oblique rows of pleopod of the male. Bonnelly de Calventi small scales or spines on flexor surface. (1974b) included a photograph of a specimen in MALE (Maricao River, Puerto Rico).—Ros- lateral view and another of the cephalic region of trum (Figure 33a,d) with margins tapering from the body in dorsal view; drawings were presented base, bearing only slightest angle delimiting pos- of the rostrum, appendices masculina and in- terior end of acumen; apex of latter slightly ov- terna, and of the antennular peduncles. Hunte erreaching distal extremity of basal segment of (1975) depicted dorsal and lateral views of the antennular peduncle; dorsal median carina cephalic region, the third, fourth, and fifth pe- gently curved, not excavate dorsally (not dipping reiopods, and a dorsal view and appendages of below level of lateral carinae posterior to acu- the first larval stage. Holthuis (1977) illustrated men), and reaching apex of acumen; ventral car- the fifth pereiopod of an ovigerous female from ina with 2 preapical teeth, both well forward on Cuba in which the chela-like extremity is shown. acumen; ocellar beak obscured between eyes and DIAGNOSIS.—Cephalic region of carapace not reaching little beyond level of base of stylocerite, conspicuously sculptured, glabrous, lacking its cephalic border sloping posterodorsally and 84 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 33.—Atya lanipes (all from male from Rio Maricao, Puerto Rico, except g,h,i from Rio Mameyes, Puerto Rico, and n from Rock Springs, Jamaica): a, dorsal view of cephalic region; b, dorsal view of antennular peduncle; c, mesial view of appendices masculina and interna; d, lateral view of cephalic region; ej, flexor surface of distal part of fourth and fifth pereiopods, respectively; g, flexor surface of distal part of third pereiopod; A,i, enlargements of distal and proximal parts of g;j, dorsal view of telson; k,l, enlargements of distal and proximal parts of m; m/i, flexor surface of distal part of third pereiopod; o, lateral view of third pereiopod; p, lateral view of second through fifth abdominal pleura; q, lateral view of preanal carina. (Scales marked in 1 mm increments.) NUMBER 364 85 dorsal border embraced by sides of ventral rostral Third maxilliped overreaching antennular pe- groove. Antennal and pterygostomian spines duncle by distal fifth of ultimate podomere of strongly acute; no spine present between them. endopod; tip of exopod attaining base of distal Surface of carapace with very fine, dense punc- fifth of penultimate podomere. tations but devoid of ridges and spines other than First pereiopod reaching level of distal extrem- those just mentioned; cephalothorax without con- ity of antennular peduncle, second extending to spicuous setae. base of distal fifth of fingers of first pereiopod; Pleura of first 3 abdominal segments (Figure terminal brush of both appendages containing 33p) with rounded posteroventral extremities; setae with scraping denticles (Figure 4b). Third corresponding parts of fourth and fifth angular pereiopod (Figure 33k,l,m,o) with lateral disto- but not produced in spines. All pleura lacking ventral spine on merus and carpus, ventral spine corneous denticles on ventral margin, but third on right merus, carpus with 2 smaller distolateral through fifth with moderately prominent fringe spines, and when appendage extended anteriorly of plumose setae. Fourth abdominal tergum overreaching antennular peduncle by dactyl and about 1.1 times as long as fifth, subequal in length four-fifths of propodus; merus with ventromesial to sixth, and almost 0.9 as long as telson. Sternum margin bowed, approximately 6 times as long as of fifth abdominal segment with very small, al- high, about 2.2 times as long as carpus, and 2.2 most rudimentary median tubercle (Figure la). to 2.4 times as long as propodus; latter about 9 Sternum of sixth abdominal segment about 0.9 times as long as wide and about 1.1 times as long as long as broad. Free part of preanal carina as carpus; distoventral margin of coxa entire (Figure 33^) acute and rather short, falling far (evenly rounded), and mesial caudoventral prom- short of caudal margin of basal part of sclerite. inence absent. Lateral, dorsal, and ventral sur- Telson (Figure 33;) little less than 2.4 times as faces of merus studded with many minute api- long as wide, its dorsal surface bearing paired cally sclerotized tubercles, most of which ar- concave rows of 6 corneous denticles, and poster- ranged in sublinear series; setae inconspicuous; omedian tubercle slightly overhanging caudal mesial extremity of podomere very weakly pro- margin. duced at level of mesial articular condyle of car- Proximal podomere of antennule (Figure 33b) pus. Carpus more strongly tuberculate, except on whh stylocerite reaching base of distal third or flexor surface, than merus, and clusters of mod- fourth of segment; dorsal surface with linear clus- erately conspicuous tufts of long plumose setae ter of setae but lacking corneous spinules; distal present ventrolaterally, most abundant near large margin bearing row of 9 (right) or 8 (left) cor- ventrolateral spine. Propodus with most tubercles neous denticles; penultimate segment of peduncle arranged in series on dorsal, mesial, and lateral little more than twice as long as wide and bearing surfaces, those on flexor surface less obviously so, 10 spinules, most arranged in linear series, on especially proximally; prominent tufts of setae dorsal surface and 10 on distal margin; ultimate present along proximal half of lateral surface. podomere with row of 8 (right) or 10 (left) spi- Dactyl movable, its flexor surface with 2 longi- nules at base of lateral flagellum and another of tudinal rows of 4 to 6 denticles each flanked 6 (right) or 5 (left) at mesial base of mesial distally by setal clusters. flagellum, none present proximal to rows. An- Fourth pereiopod with dactyl reaching slightly tenna with ventrolateral spine on basis not reach- beyond distal margin of carpus; length of merus ing so far anteriorly as stylocerite; lateral spine almost twice that of carpus, latter approximately on scaphocerite rather strong, reaching base of 0.9 as long as propodus. Fifth pereiopod reaching middle third of ultimate podomere of antennular anteriorly as far as fourth; merus almost 1.7 times peduncle; lamella far surpassing latter; flagellum as long as carpus, latter 0.6 as long as propodus. extending almost to midlength of telson. Ornamentation of merus, carpus, and propodus 86 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY of fourth pereiopod similar to that of third except Holthuis (1963:61) to be 25 (73) and 23 (70) mm, for 3, instead of 2, distolateral spines on carpus. respectively. Carapace lengths of specimens from Ornamentation of fifth pereiopod like third and Puerto Rico reported by Chace and Hobbs fourth but with 2 ventral spines on merus proxi- (1969:62) for 16 males ranged from 4.2 to 28 mm, mal to distal ventrolateral spine, and row of 5 for 6 females 4.9 to 21.8 mm, and for 9 ovigerous distolateral spines on carpus. females, 9.4 to 17.7 mm. The carapace lengths of Diaresis of lateral ramus of uropod flanked Hunte's (1975:66) four females from Jamaica proximally by row of 22 (right) or 23 (left) artic- were 9.8, 12.6, 13.7, and 14.0 mm. The specimen ulated corneous denticles, and fixed spine at lat- from the Jamaican cave recorded by Peck (1975) eral end of row. is a male having a carapace length of 21.5 mm. COLOR NOTES.—From Hunte (1975:67). Villamil and Clements (1976:21) measured 73 specimens: the 34 males had total lengths of 50.25 The ground colour of the cephalothorax and abdomen is to 74.55 (average (61.87) mm; the non-ovigerous dark green, but the presence of scattered cream chromato- phores laterally produces a mottled effect. The cream chro- females, 30.15 to 62.90 (average 44.18) mm; and matophores tend to become white in a dorsovent ral direction the single ovigerous female, 34.25 mm. on the lateral surfaces, and along the ventral margins of the DISTRIBUTION AND SPECIMENS EXAMINED.—This pleura form a distinct white line. There is a mustard-col- shrimp has been collected on only six islands of oured strip running centrally on the dorsal surface from the tip of the rostrum to the tip of the telson where it terminates the West Indies: Cuba, Jamaica, Hispaniola, in an orange-coloured chromatophore. The strip remains Puerto Rico, Saint Croix, and Saint Thomas narrow for the length of the carapace, but then widens to (Figures 34, 35). form a series of triangles, one in each abdominal segment. In Records for the known localities are listed be- the first two segments the apex of the triangle is posteriorly low. Collections that we have examined are directed, in the last three anteriorly directed. In the last two segments the strip covers the bulk of the dorsal surface. The marked with an asterisk if they have been previ- remainder of the dorsal surface of the abdomen, i.e. the area ously reported and with a dagger if they are on either side of the strip, is dark green; as is the posterior reported herein for the firsttime . Numbers follow- half of the dorsal part of the carapace. The anterior half of ing the specimens listed are measurements in mm, the dorsal surface of the carapace is tan-coloured on either side of the strip. of the carapace length. Some listings lack the date the collection was made and/or the name of the The antennular and antennal peduncles are mottled green; the flagella are tan. The antennal scales are faintly collector; these could not be determined. mottled green at the base, but this quickly grades into CUBA: (1) RMNH, Rio Caburny near Topes de translucency. All five pairs of pereiopods are distinctly Collantes, Provincia de Las Villas (Holthuis, banded with alternating strips of white and olive green. The 1977:271), 1 ovig$ (12.0), "Cubano- Roumanian tufts of hair borne distally on pereiopods 1 and 2 are Biospeleological Expd to Cuba." (2) fUSNM, strikingly orange in colour. All pleopods are white at their base, quickly belonging translucent distally. Trinidad Mts, Provincia de Las Villas, 1 ovig 9 (11.0), 5 Jun 1959, M. Westfall. (3) fUSNM, Bonnelly de Calventi (1974b) found the color- Baracoa, Provincia de Oriente, 16* (6.2), 30 Jan ation of this shrimp in the Dominican Republic 1902, W. Palmer. (Holthuis, 1977:272, reported to be quite variable: dark brown, or sometimes a juvenile collected in Rio Ceiba with a carapace light with irregular cream spots over the entire length of 5.5 mm that he suspected as being a body. member of this species.) SIZE.—The largest specimen we have examined JAMAICA: (1) Cane River, Saint Thomas Parish is a male from Puerto Rico having a carapace (Hunte, 1975:67), 4$. (2) fUSNM, Rock Springs length of 29.0 mm; that of the largest female, Cave, 0.9 km E of Pear Tree Grove, alt 400 m, from Saint Croix, 18.2 mm. The carapace length Saint Mary Parish (Peck, 1975:308, locality not and total length (latter in parentheses) of the cited), 16 (21.5), Aug 1974, S.B. Peck. (3) Wag- holotypic and paratypic males were reported by water rivulet N of Kingston, alt 700 m, Saint NUMBER 364 87

FIGURE 34.—Distribution of Atya lanipes (circled numerals = number of localities; see Figure 35).

Andrew Parish, 1 specimen (fide L.B. Holthuis). 9 (17.5), spring 1954, L.A. Costas Grava. (4) PUERTO RICO: (1) *USNM, Rio Culebrinas at *USNM, Rio Lajas, 2800 m E and 3,000 m S of Rte 13, 1,000 m S and 300 m E of San Sebastian Vega Alta (Chace and Hobbs, 1969:62), 16 (4.6), (Chace and Hobbs, 1969:62), \6 (16.2), 2 Jun 19 (9.3), 8 May 1953, HWH. (5) *USNM, Rio 1953, H.W. Harry. (2) *USNM, Rio Maricao at Cibuco at Rte 20, 1,500 m S and 3,500 m W of Maricao (Chace and Hobbs, 1969:62), 26 (21.0, Corozal (Chace and Hobbs, 1969:62), 56 (5.7- 22.7), 2$ (13.2, 13.4), 1 Feb 1965, N.T. Mattox. 11.1), 2$ (5.2, 5.9), 2 ovig 9 (9.6, 11.8), 5 May (3) *USNM, freshwater streams at Jayuya (Chace 1953, HWH. (6) *USNM, Rio Manti at rd from and Hobbs, 1969:62), 1

FIGURE 35.—Distribution of Atya lanipes in the Greater Antilles.

\6 (23.4), 2 ovig 9 (14.9, 18.0), 23 Nov 1954, Hobbs, 1969:62), 16 (about 6.2), 19 Feb 1899. (9) HWH. (7) *USNM, trib to Rio Mameves at Rte fUSNM, below dam on Rio Cubuy at Rte 112, 112, 100 m N and 1,500 m E of El Yunque 500 m S and 1,000 m W of El Yunque summit, summit (Chace and Hobbs, 1969:62), 36* (19.6- 3$, (5.5-12.9), 8 Jun 1953, HWH. (10) fUSNM, 20.5), 19 (15.5), 2 ovig $ (14.1, 14.9), 8 Jun 1953, creek below Maricao, 16* (5.8), 19 (5.3), 7 Feb HWH. (8) *USNM, El Yunque (Chace and 1934. (11) fUSNM, falls of Ri'o Grande de Ai- NUMBER 364 89 bonito, \6 (17.1), 1$ (11.5), 26 Jan 1899, Fish them. The longitudinal rows of spinules on the Hawk. (12) fUSNM, Luquillo Nat Forest, 36 dorsal surface of the telson range from five to (15.0-20.2), 9 Mar 1934, S.F. Hildebrand. (13) eight, with differences of as many as two in the *USNM, San Juan Market, 16 (16.6-18.5), 1$ paired rows on an individual specimen. (18.0), 14 Jan 1899, Fish Hawk. (14) almost all Holthuis (1977) cited several differences be- parts of the upper Rio Espiritu Santo Basin at tween his ovigerous female and the types, includ- elevations of 550 to 820 m (Villamil and Clem- ing the less conspicuous pubescence on the legs, ents, 1976:15-16, 22). (15) "at low elevations in the presence of an "anteroventral spine" on the other forested areas of the Espiritu Santo River carpus of the third through fifth pereiopods and basin " (Villamil and Clements, 1976:40). on the merus of the fourth and fifth. Spinules are HISPANIOLA: (1) Canada Madrigal, Dominican distributed over the surface of the carpus and Republic (Bonnelly de Calventi et al., 1973:1338; propodus, and on the fifth pereiopod the long Bonnelly de Calventi, 1974a: 16), 16* (18.9), 79 ventrodistal spine forms a "chela-like structure" (11.7-17.2). (2) Canada de los Anones, Municipio with the dactyl. de Villa Altagracia, Dominican Republic (Bon- Among the specimens examined by us, the nelly de Calventi, 1974b:39). rostrum shows marked variation in shape; in some SAINT CROIX: (1) fUSNM, Fairplain Street, Id* of the smaller individuals the margins are some- (5.4), 1 ovig 9 (12.5), H.A. Beatty. (2) fUSNM, what suddenly contracted some distance posterior Caledonia Street, 46 (16.9-18.2), 29 (9.1, 13.2), 4 to the apex (Figure 36), but the angles are lacking; ovig 9, (16.9-18.2), HAB. (3) fUSNM, Chris- the margins in most specimens, however, taper tiansted, 19 (7.3), HAB. (4) *USNM, no locality from the base; the apex may not reach the distal data, 4 juv (3.2-3.3), 22 Mar 1937, HAB. (5) extremity of the first segment of the antennular fUSNM, Cave at Juan Diaz, 29 (13.0, 17.1), 21 peduncle, or it may extend as far as the end of Feb 1932, Chapman Grant. the proximal third of the penultimate segment; SAINT THOMAS: *RMNH, no locality data (hol- the ventral keel may lack teeth or there may be otype and paratype), 26 (18.1, 19.9). as many as three. Most of the specimens exhibit VARIATIONS.—Chace and Hobbs (1969:63) paired rows of six spinules on the dorsal surface pointed out that in none of the specimens from of the telson, but there may be as few as five and Puerto Rico examined by them are "the last three as many as eight in at least one of the two rows. pereiopods clothed in hair dense enough to con- The variations noted by the above mentioned ceal the underlying surface." They also noted authors have been observed by us. As for the considerable variation in the pterygostomian an- distal part of the fifth pereiopod, there is consid- gle; it is longer and more slender in the larger erable variability in the size of the distoventral males than in the types but less prominent in spines on the propodus, and although the chela- smaller specimens of both sexes, "often being like arrangement mentioned by Holthuis has reduced to no more than a broadly acute angle." been observed, some of our specimens have spines Hunte (1975:67, 68) in discussing the "Pecu- opposing the dactyl that are very much reduced, liarities of, and variation within, Jamaican spec- others have paired long ones, but in most in imens" noted that the rostra attain the proximal which the fifth leg remains intact the spines are one-fourth to one-third of the second article of intermediate in size; furthermore the long spines the antennular peduncle, and one or two teeth occur in males as well as in females. are present on the ventral surface. Minor varia- ECOLOGICAL NOTES.—All of the habitats cited tions are pointed out in the spination of the by Chace and Hobbs (1969:62) in Puerto Rico podomeres of the third, fourth, and fifth pereio- were lotic, but the data recorded by them con- pods, and, like Chace and Hobbs, he noted con- tained no further ecological notes. Bonnelly de siderable variation in the pubescence borne on Calventi (1974b:38) observed that in the Domin- 90 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 36.—Atya lanipes, variation in the rostrum of 3 specimens from Rio Manti, Puerto Rico (numbers in parentheses * carapace length in mm): a, ovigerous female (14.9); bt ovigerous female (18.0); c, male (23.4). ican Republic this shrimp frequents tranquil, that at one of their stations a shrimp would filter shaded waters beneath the "hojarascas" floating 0.00006 gm of paniculate matter per hour (p. on the surface, but it also lives beneath stones. 48). They also observed that "sex exclusion in Hunte (1975:67) stated that his specimens "were habitat preference" existed throughout the study collected in pure freshwater where the rate of area (p. 29). Males occupied the parts of pools water flow was high and bottom substrate stony." where the current was strongest, and boulders Peck (1975:308) reported the presence of Atya were a prominent feature of the substrate. In lanipes in a Jamaican cave where he collected it contrast, females tended to frequent the marginal in company with Atya innocous and Macrobrachium areas where the flow was weaker and where the species beyond the entrance pool, "deeper in the substrate consisted of rubble and gravel. cave. The water temperature was 23.3° C ... . LIFE HISTORY NOTES.—The ovigerous females Tubificid worms and psychodid fly larvae were reported by Chace and Hobbs (1969:62) to have abundant in guano pools" (p. 320). Villamil and carapace lengths of 9.4 to 17.7 mm were collected Clements (1976) found this shrimp in the Espiritu on 8 June 1953, 23 November 1954, and 5 May Santo Basin, Puerto Rico, in every stream where 1955. The oval eggs ranged from 0.3 X 0.6 mm to shrimp were present and "where apparently A. 0.4 X 0.7 mm. The latter range was reported by innocous cannot live" (p. 4). They concluded that Holthuis (1977:271) for the ovigerous female col- it was not restricted in the streams by any physical lected in Cuba. Bonnelly de Calventi (1974b:39) or chemical characteristic investigated (p. 26). recorded the occurrence of ovigerous females in According to them, Gifford and Cole (1970, un- February, March, April, May, and August. published) believed Atya lanipes to be Hunte (1975:69-72) described and illustrated the "first stage larvae" that hatched from eggs borne the single most important specie[s] for the recycling of detritus and nutrients washed into the stream .... In a pool by a female having a carapace length of 9.8 mm. approximately 48 square feet, a minimum of 26 individuals These larvae were planktonic and swam con- were counted by Gifford and Cole. Through skin-diving and stantly, tail first and on their backs. They were [sic] average of 25 individuals per square meter were counted attracted to light and were not apparently over large boulders in riffle areas. [Villamil and Clements, harmed by a change of salinity from 0 to 20%o 1976:28, 29] over a period of three hours. They failed to accept Both filter feeding and direct grazing were ob- any food offered them. The nearly transparent served. Villamil and Clements (1976) estimated body, 1.73 to 1.78 mm in length and bearing red NUMBER 364 91 chromatophores, was slightly bent at the third REVIEW OF LITERATURE.—Atya margaritacea was abdominal segment. Illustrations of the dorsal described by A. Milne-Edwards (1864) who based view of the larva and its appendages were in- his account on specimens that he believed to have cluded. Neither the date of collection of the ovi- been collected on New Caledonia. Smith (1871) gerous female nor the diameters of the eggs was described the same species from Nicaraguan spec- recorded. Villamil and Clements (1976:40) stated imens, designating them Atya rivalis. In describing that the only ovigerous female they captured was Atya gabonensis, Giebel (1875) noted the previously obtained in November, but others were observed known members of the genus and included a in the Espiritu Santo River basin from December reference to Milne-Edwards' species. Kingsley to February. They concluded that the breeding (1878a) contrasted certain characters of A. rivalis season extends from November to March with with those of his Atya punctata and mentioned the larval release occurring in February. The new species in a second publication that year (1878b); and previously recorded data summarized here in the latter he also recorded A. scabra from fresh- indicate that egg laying occurs, somewhere within waters of western Mexico, a record almost cer- the range of the species, from November to Feb- tainly based upon a misidentification of A. mar- ruary and in May and June. garitacea or A. ortmannioides. According to Pocock (1889), "the features pointed out by Mr. Kingsley to distinguish his punctata from ... rivalis ... are . .. regarded as due either to individual vari- Atya margaritacea A. Milne-Edwards ation or to difference of age." Bate (1888) did not FIGURES 1*. 9,10,1 ld,e, 37-41 add to our knowledge of the species. Two years later, however, Ortmann (1890) presented a brief Atya margaritacea A. Milne-Edwards, 1864:148, 149, pi. 3, fig. diagnosis and illustrated the rostrum of A. mar- 2 [type-locality: New Caledonia (as pointed out by Hol- thuis, 1966:234, "evidently in error"); types: MHNP, No. garitacea and placed it, together with A. robusta, 601, 26, 1$].—Giebel, 1875:52.—Bate, 1888:693.—Ort- A. scabra, and A. sculptipes (= sculptata), in his mann, 1890:465, 466, pi. 36: fig. 7; 1895:408, 409.— "margaritacea-Gruppe." In his study of the family Bouvier, 1905:121, 122.—Holthuis, 1969:92. Atyidae, Ortmann (1895:408) included A. margar- Atya rivalis Smith, 1871:94 [type-locality: "Fresh water itacea in his key to the members of the genus, but streams, Polvon [12°21'N, 87°05'W], Occidental Depart- ment, Nicaragua"; types: MCZ 317, 3 dismembered 6 in he expressed the opinion that "the locality given one bottle, single specimen of undetermined sex in an- by Milne-Edwards for margaritacea and robusta is other].— Kingsley, 1878a:92; 1878b:57.—Pocock, 1889: not correct.. .." Among the several synonyms he 16.—Oliveira, 1945:179.—Holthuis, 1966:234.—Chace cited for A. scabra was Atya rivalis, and the same and Hobbs, 1969:66.—Solar, 1972:8.—Abele, 1975:56, recognition was accorded these two species by 57.—Abele and Blum, 1977:240-242, 245, 250.—Mendez, him in his study of the shrimps of South America 1981:14, 70, pi. 4: fig. 34; pi. 27: figs. 209-211.—Rodri- guez, 1981:46. (1897). Rathbun (1900, 1901) seems to have Atya scabra.—(?)Kingsley, 1878b:57.—Ortmann, 1895:408- agreed with Ortmann that A. rivalis is a synonym 410, 415, 416; 1897:183-185.—(?)Rathbun, 1900:313; of A. scabra, for she listed Nicaragua in citing the 1901:119.—Bouvier, 1904:138; 1905:112, 121, fig. 25M; distribution of the latter, and, to our knowledge, 1925:293, 314-317, figs. 703, 704.—De Man, 1925:28.—J. no locality record for A. scabra in Nicaragua exists. Roux, 1926b:217.—Monod, 1933:461.—Schmitt, 1935: 136.—Coventry, 1944:534.—(?)Oliveira, 1945:177.—Hol- Even though Rathbun did not cite A. rivalis thuis, 1951:25; 1966:234, 238; 1969:92.—Parodiz, among the synonyms of A. scabra, we suspect that 1960:38, 39 [only specimens from Costa Rica, not figs. 1- her inclusion of that country in the range of A. 3].—Chace and Hobbs, 1969:63. [All in part.] scabra was based upon Smith's record for Atya Atya scabra var. margaritacea.—Monod, 1967:140, pi. ix: figs. rivalis, perhaps influenced by Ortmann. 13, 14 [not 15 and 16 as noted by Monod; see comment Bouvier (1904) cited Atya margaritacea as a syn- under "Published Illustrations" below]. Ataya margaritacea.—Chace and Hobbs, 1969:63 [erroneous onym of A. scabra and the following year (1905) spelling]. added A. rivalis, among others, to his list of syn- 92 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

onyms of A. scabra, and both A. margaritacea and roneous data. Holthuis (1969) and Chace and A. rivalis were included among the synonyms of Hobbs (1969) included A. margaritacea in their A. scabra in his monograph (1925). De Man (1925) synonomy of A. scabra, but the latter tentatively also considered A. rivalis to be a synonym of A. regarded A. rivalis as a distinct species. The latter scabra. Atya margaritacea was relegated to the syn- opinion was concurred in by Abele (1975). Solar onomy of the latter by J. Roux (1926b), but he (1972) included A. rivalis in his catalogue of the expressed doubt that his shrimp occurs on New Crustacea of Peru, citing Holthuis (1966) and Caledonia, for among 150 atyids collected in six Chace's identification of specimens from rios Chi- localities there, no representative of the genus cama and Jequetepeque. In their study of the Atya (as restricted herein) was among them. freshwater decapods of the Perlas Archipelago, Monod (1933) and Schmitt (1935) followed Bou- Panama, Abele and Blum (1977) found A. rivalis vier in citing New Caledonia within the range of only on San Jose; their six specimens ranged in A. scabra. Nicaragua was included among the length from (36*) 40.4 to 32.4 mm and (39) 6.5 to areas frequented by A. scabra by Oliveira (1945), 10.4 mm. A summary of their observations on the an inclusion that almost certainly was based upon habitat occupied by this shrimp on the island is the assumed identity of A. scabra and A. rivalis. included in "Ecological Notes" below. Mendez Coventry (1944) cited a new locality in Panama (1981) reported the occurrence of A. rivalis in six for a shrimp that was identified by him as A. river basins in Peru between 3°47'S and 1 l°04'S scabra. In all probability this record is for A. and at altitudes of 0 to 210 m. Illustrations are margaritacea. also provided. Although Holthuis (1951) did not list A. rivalis PUBLISHED ILLUSTRATIONS.—In the original de- among the synonyms of A. scabra, he reported the scription of the species, A. Milne-Edwards (1864) species' occurrence on the west coast of America included a lateral view of the entire animal, dorsal from Lower California to Costa Rica, an area and lateral views of the rostrum, and a dorsal inhabited by A. margaritacea instead of A. scabra. view of the telson. The only other figures avail- He also noted that the latter had been reported able are those of Ortmann (1890), a dorsal view from New Caledonia, no doubt referring to of the rostrum, and of Bouvier (1905), a dorsal Milne-Edwards' locality. Parodiz (I960) reported view of the rostrum and lateral view of the ce- the presence of this shrimp (identified by him as phalic region of the carapace. The last two were Atya scabra) from Rio de Colon, Costa Rica. In reprinted but mislabeled by Bouvier (1925), for his study of the freshwater shrimps of Annobon, figures 703 and 704 correspond to those labeled Holthuis (1966) pointed out certain differences as "A margaritacea" in 1905, fig. 25 M. Inasmuch between A. scabra and specimens of Atya from as Monod's (1967) figures were taken from Bou- Ecuador and Peru. Those specimens from the two vier (1925), those attributed to "Typical" scabra South American countries (which have relatively and to the "var. margaritacea" are reversed. Men- shorter and broader segments of the antennular dez (1981) included two drawings of Atya rivalis peduncle, the second segment being much longer in lateral view, one of a dorsal view of the rostrum, than the third, and in which the mesiodistal lobe and another of the first pereiopod. of the merus of the third pereiopod forms a strong DIAGNOSIS.—Cephalic region of carapace not "bluntly triangular process," and the legs more conspicuously sculptured; spines limited to anten- robust than in the eastern American and West nal, moderately strong pterygostomian, and oc- Indian A. scabra) he assigned to A. rivalis. In casional ventral rostral. Rostrum with margins commenting on the range of A. scabra recorded by suddenly contracted forming distinct angle ante- Bouvier (1925), he stated that records from New rior to orbit, angle often produced anteriorly; Caledonia are almost certainly based on speci- dorsal surface without median row of strong mens (= syntypes of A. margaritacea) bearing er- spines. Ventral margin of third through fifth NUMBER 364 93 abdominal pleura with short to moderately long ygostomian angle produced in short acute spine; rows of corneous denticles, such never present on no spines present between antennal and ptery- second; caudoventral angle of fourth and fifth gostomian spines. Carapace (length 24.3 mm) pleura not produced in spines. Sternum of fifth devoid of other spines and conspicuous ridges; abdominal segment with comparatively large me- surface studded with short erect setae borne in dian tubercle, that of sixth about 1.6 times as rather crowded punctations. broad as long. Preanal carina with short, some- Pleuron of second abdominal segment (Figure what compressed spine directed caudoventrally, 37*) with rounded posteroventral extremity, that spine not reaching level of caudal margin of basal of third, fourth, and fifth angular although none part of sclerite bearing it. Telson 1.6 to 1.8 times produced in spines. Ventral margin of third, as long as broad with 5 to 7 spines in each of 2 fourth, and fifth pleura with linear clusters of 15, dorsal rows. Antennular peduncle with dorsal 17, and 9 corneous denticles, respectively; that of surface of proximal article usually devoid of scle- sixth with sparse setal fringe posteriorly. Fourth rotized spinules proximal to distal transverse row; abdominal tergum approximately 1.2, 1.2, and penultimate article 1.4 to 1.6 times as long as 1.1 times as long as fifth, sixth, and telson, re- wide, and dorsal surface with many scattered spectively; length of sixth subequal to that of spinules. Coxae of third and fourth pereiopods fifth, and 0.9 as long as telson. Sternum of fifth lacking prominent anterolateral spine. Third pe- abdominal segment (Figure li, 37^) with rather reiopod with merus rounded ventrally, 2.1 to 2.4 large, compressed median tubercle; sternum of times as long as high, ventromesial surface slightly sixth 1.6 times as broad as long. Preanal carina to strikingly bowed, never parallel to that of (Figure 37/) with short spine, tip of which barely corresponding podomere of other member of pair, reaching midlength of sclerite. Telson (Figure and lateral surface bearing corneous spines and 37m) about 1.8 times as long as wide, its dorsal tubercles, many of latter with flattened scalelike surface bearing paired mesially concave rows of extremities, most arranged in linear series; pro- 6 denticles each and posteromedian tubercle, lat- podus 1.7 to 1.9 times as long as broad, studded ter slightly overhanging midcaudal margin. with scalelike tubercles on extensor surface and Proximal podomere of antennule (Figure 37b) with moderate number of similar, rather large with strong, acute, corneous-tipped stylocerite ones scattered on flexor surface, if any of those on overreaching midlength of segment, dorsal sur- latter arranged in row, those comprising it never face with linear cluster of setae, and left member contiguous or overlapping; dactyl not fused with with 1 (right with 2) premarginal cornified spi- propodus and bearing single distinct straight to nules, distal margin studded with dorsolateral slightly curved, longitudinal row of scalelike tu- row of 7 cornified spinules; penultimate segment bercles on flexor surface. of peduncle subequal in length and width, its SYNTYPIC MALE.—Eyes collapsed but obviously dorsal surface bearing 15 (right) or 13 (left) small, well developed and presumably pigmented. Ros- corneous spinules, more laterally situated ones trum (Figure 37a,d) with slightly concave margins forming curved row, dorsodistal margin with row produced in acute angles flanking base of acu- of 6; ultimate podomere of peduncle, little more men, latter slightly overreaching basal podomere than half as long as penultimate segment, armed of antennule; high, rounded dorsal carina exca- with rows of 8 similar spinules at dorsal base of vate dorsally (dipping below level of lateral cari- each flagellum and groups of 3 or 4 proximal to nae posterior to acumen) and almost reaching tip row flanking base of lateral flagellum. Antenna of acumen. Ventral carina with 1 small tooth with ventrolateral spine on basis almost reaching near apex of acumen; ocellar beak terminating in level of tip of stylocerite; lateral spine on scapho- acute angle, dorsal margin horizontal and ante- cerite extending to about midlength of ultimate rior margin oblique. Antennal spine acute; pter- podomere of antennular peduncle, falling short 94 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 37.—Atya margaritacea (all from syntypic male): a, dorsal view of cephalic region; b, dorsal view of antennular peduncle; c, mesial view of appendices masculina and interna; d, lateral view of cephalic region; e, lateral view of second through fifth abdominal pleura; /, flexor surface of distal pan of right third pereiopod; g, preanal carina, sternum of sixth abdominal segment, and median tubercle of fifth; h,i, flexor surface of distal part of fifth and fourth pereiopods, respectively;^, lateral view of preanal carina; k, flexor surface of distal part of left third pereiopod; /, lateral view of third pereiopod; m, dorsal view of telson. (Scales marked in 1 mm increments.) NUMBER 364 95 of distal end of peduncle of antenna; lamella of face with single longitudinal row of 5 denticles scaphocerite overreaching peduncles of both an- flanked distally by paired setal clusters. tennule and antenna. Flagellum of antenna bro- Fourth pereiopod with dactyl reaching end of ken. proximal fourth of carpus of third; length of Third maxilliped overreaching antennular pe- merus 2.3 times as long as carpus, latter slightly duncle by about 0.25 length of ultimate podo- shorter than propodus. Fifth pereiopod (lacking mere of endopod; tip of exopod reaching base of propodus and dactyl) with merus about 1.4 times apical third of penultimate podomere, latter as long as carpus. Ornamentation of merus, car- about 1.6 times as long as ultimate. pus, and propodus of fourth pereiopod consisting First pereiopod reaching base of distal third of of strong distoventral spine on both merus and ultimate podomere of antennular peduncle; sec- carpus, smaller ventral spine at base of distal ond pereiopod attaining distal fourth of fingers third on merus, and row of 3 corneous tubercles of first; terminal brush of both appendages ap- on distolateral margin of carpus; distal ventrolat- parently lacking scraping denticles. Third pereio- eral part of merus, entire ventrolateral part of pod (Figure 37/,/:,/) lacking spines, and, when carpus, and proximolateral two-thirds of propo- extended anteriorly, overreaching antennular pe- dus with conspicuous tufts of long plumose setae. duncle by length of distal three-fourths of carpus Ornamentation of merus and carpus of fifth pe- and entire propodus and dactyl. Merus with ven- reiopods similar to that of fourth except merus of tromesial margin bowed, 2.6 times as long as right with 2 ventral spines and 4 instead of 3 high, twice as long as carpus, and 2.4 times as distolateral tubercles on carpus. long as propodus; latter 1.8 times as long as wide Diaresis of lateral ramus of uropod flanked and 0.9 as long as carpus; distoventral margin of proximally by row of 16 articulated corneous coxa with prominent tubercle, otherwise gently denticles and fixed spine at lateral end of row. curved (not scalloped), and mesial caudoventral COLOR NOTES.—The following remarks con- prominence setiferous but very weak. Lateral, cerning the color of this shrimp are based on dorsal, and ventral surfaces of merus studded transparencies of specimens from £1 Valle, Pan- with linear series of large tubercles bearing cor- ama, lent by Bruce E. Felgenhauer and on obser- nified discs, free margins of which sharp; each vations he kindly communicated to us. He has tubercle flanked by semicircular cluster of con- observed what he terms three "fairly distinct spicuous plumose setae; longitudinal row of long phases": in juveniles the ground color is light setae on lateral surface made inconspicuous by brown and there are two or three transverse dark setae flanking subtending tubercles; mesial ex- bands on the abdomen; larger individuals are tremity of podomere produced in moderately dark brown, resembling adult A. tenella {— A. strong lobe at level of mesial articular condyle of innocous) and the largest become uniformly blue- carpus. Carpus strongly tuberculate, tubercles black. tipped with corneous discs, latter increasingly In the transparencies of an individual with a produced toward distal end of podomere, tuber- carapace length of eight mm, the only conspicu- cles flanked distally by setae as on merus, setae ous markings are two dark transverse bands, one long and forming conspicuous tufts ventrolat- spanning the posterior margin of the carapace erally. Propodus also strongly tuberculate with and anterior part of the first abdominal segment, corneous apices ranging from scalelike proximally and the other extending across the anterior half to clawlike distally, those on flexor surface some- of the sixth abdominal segment. A few dark what irregularly dispersed (none arranged in con- patches occur dorsally, one pair situated lateral tiguous or subcontiguous linear series); clusters of to the base of the rostrum, and another in the plumose setae on lateral surface conspicuous at posterior gastric region. least proximally. Dactyl movable, its flexor sur- The photograph of an animal with a 20 mm 96 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY carapace length agrees very well with the illustra- tion of A. scabra herein (Figure 47) except that the dark distal part of the third pereiopod is not so strikingly conspicuous, and there are a few dark spots on the lateral surface of the carapace that are not present in the illustration of A. scabra. Whether or not there is a dorsomedian longitu- dinal stripe cannot be determined from the trans- parency because of the position of the animal in the photograph. The uniform, very dark-blue coloration, even extending onto the basal section of the eyestalks of an animal with a 30 mm carapace length suggests that an encrustation of some kind is at least partly responsible for the intensity of the dark color. SIZE.—Of the 98 males measured, the largest specimen has a carapace length of 38.0 mm; the average was 18.6 mm. The largest of 54 females FIGURE 38.—Distribution of Atya marganlacea (for locality has a corresponding length of 24.4 mm, the av- records in Costa Rica and Panama, see Figure 39). erage 14.3 mm. The smallest and largest oviger- ous females have carapace lengths of 7.3 and 20.8, average 12.7 mm. DISTRIBUTION AND SPECIMENS EXAMINED.—Dis- regarding the almost certain erroneous data (New Caledonia) recorded by A. Milne-Edwards for his types as well as for those reported for Argentina and , this American species ranges from Baja California southward on the Pacific versant to Peru (Figures 38, 39). Records for the known localities are listed be- low. Collections that we have examined are marked with an asterisk if they have been previ- ously reported and with a dagger if they are reported herein for the firsttime . Numbers follow- ing the specimens listed are measurements, in mm, of the carapace length or, if followed by "t.l.," total length. Some listings lack dates and/ or collectors; if so, these could not be determined. MEXICO: (1) fUSNM, La Paz, Baja California Sur, 1$ (9.6), Jun 1882, L. Belding. (2) fMCZ, FIGURE 39.—Distribution of Atya margarilacea in Costa Rica San Jose, Baja California Sur, 26 (6.1, 8.2), 2$ and Panama. (7.5, 7.9), 3 juv, Gustav Eisen. (3) fBM, San Ramon, 36* (28.8-35.6), 1$ (22.7), A.C. Butler. (4) (14.8-19.3), F.D. Godman. (6) fBM, Rfo Ma- fUSNM, Rio San Nicholas between Chamela chiro, trib to Rio Grande, near San Mateo, Oax- and Tomatlan, Jalisco, 86* (6.1-12.0), 2$ (7.8- aca, alt 300 m, 26 (27.9, 28.3), H. Pittier de 9.9), 4 ovig $ (7.3-8.4), 1 juv (3.1), 6 Apr 1965, Fahega. R.R. Miller and Sable. (5) fBM, San Isidro, 36* GUATEMALA: (1) fBM, Pentalion, 106* (13.1- NUMBER 364 97

22.5), 109 (9.1-16.8), Champion. (2) fUSNM, Heck. (13) fLGA, first stream on Carti Rd, km Rio La Paz at San Diego on Carretera Intera- 11, 26* (15.3, 18.3), 15 Jan 1975, DK. (14) fLGA, mericana, 20+ km W of Jutiapa, 46* (14.6-30.8), second stream on rd to Chepo, 36* (7.0-31.2), 19 13 Mar 1946, R.R. Miller and Holloway. (approx 8.5), May 1978, LGA, BEF, LD. Provincia EL SALVADOR: fUSNM, Santo Tomas in de Los Santos—(15) fUSNM, Rio Tonosi, halfway flooded river, 76* (11.8-15.0), 1$ (8.4), 13 Mar between Llano de Piedra and Tonos Plain, 16* 1939, Calderon. (16.0), 19 (14.4), Mar 1963, HL. Provincia de Ver- NICARAGUA: fMCZ, El Polvon, 4 specimens aguas—(\6) fUSNM, creek 13.6 km W of El (syntypes of Atya rivalis, in very poor condition), Maria on Sona-Remedios Rd (Tabasara Basin), 1871, J.A. McNiel. 66* (7.6-21.4), 49 (about 10.0-15.2), 2 ovig 9 (9.3, COSTA RICA: (1) fUSNM, Rio Riryto, 13.6 km 11.4), 11 Nov 1961, HL. (17) fUSNM, Rio Ca- from Rincon de Osa, 46* (10.6-17.5), 3$ (11.4- lixto, 11.9 km W of El Maria on Sona-Remedios 17.3), 1 ovig 9 (10.1), 15 Sep 1966, S.B. Peck. (2) Rd (Tabasara Basin), 16* (10.7), 11 Nov 1961, fUSNM, Rio de los Platanales, Golfo Dulce, 1 HL. (18) fUSNM, creek just before entering ovig 9 (9.5), Apr 1896, H. Pittier. Santa Fe (Santa Maria Basin), 56* (17.4-28.0), 19 PANAMA: Provincia de Chiriqui—{\) fUSNM, riv- (12.5), 9 Feb 1962, HL. (19) fLGA, Rio San ers flowing from Mt. Chiriqui, alt 426 m, 8°27'N, Pablo, 36* (20.4-23.2), 29 (12.1, 13.9), 30 Mar 82°24'W, 176* (20.3-30.6), 19 (20.6), 1 ovig 9 1973, LGA, P. Campanella, J.B. Graham. (20) (20.8), Jul 1883.(2) fUSNM, 17 km W of Rio fUSNM, 8.5 km E of Sona on San Remedios Rd David at Carretera Interamericana, 26 (16.6, (San Pablo Basin), 26* (13.4, 15.8), 19 (14.0), 28 23.0), 19 (24.4), 15 Apr 1962. (3) fUSNM, Rio Jan 1962. (21) fUSNM, Rio San Martin Grande, Tinta, 4.8 km W of Rio Tabasara on Sona-Re- 6.8 km S of Santiago on Montigo Rd, 36* (13.3- medios Rd (Tabasara Basin), 1 ovig 9 (9.1), 11 25.6), 14 Jan 1962, HL. (22) fUSNM, Rio Cur- Nov 1961, H. Loftin. (4) fUSNM, upper trib of vibora, 3.4 km E of Santiago on Carretera Inter- Rio Jacaque, 86* (10.2-23.4), 19 (11.6), HL and americana (San Pedro Basin), 26* (16.8, other Dean. (5)?Mt. Sapo, Pinas Bay (Coventry, injured), 21 Oct 1961, HL. (23) fLGA, Rio Las 1944:534), 1 specimen, 21 May 1941, Fifth Guias at Rd from Calobre, 26* (8.2, 28.9), 49 George Vanderbilt Expedition. Provincia de Code— (15.3-17.7), 21 Feb 1973, LGA, MHR. (24) (6) fMCZ, Rio Las Lajas, 36* (7.3-8.1), 39 (11.1- fLGA, Rio Vigui [= Rio Tabasara], 16* (26.9), 29 12.6), 3 ovig 9 (11.6-13.1), 22 Oct 1939, G.B. Mar 1973, LGA, PC, JBG. Fairchild. (7) fLGA, Rio Cocoli (Cocle?) at K-6 ECUADOR: (l)fRNHL, Chula, 36* (14.5-17.7), crossing, 26* (14.0, 21.6), 39 (9.1-12.8), 26 Feb 29 (11.1, 14.6), 3 ovig 9 (11.2, 14.1, 19.1), Sep 1973, L.G. Abele, M.H. Robinson. (8) fLGA, El 1956, W. Forster. (2) fBM, Rio Sapayo, NW Valle, Masso Charro?, \6 (13.3), 14 Apr 1973, Ecuador, 36* (23.9-24.7). LGA, L.R. Abele; 16* (22.5), 19 (19.9), 18 Apr PERU: (1) fRNHL, Rio Tumbes, 16* (24.3), 1973, LGA, LRA. (9) fLGA, El Valle in fast Mario Pena. (2) fRNHL, "don Museo Historia riffles, 1 ovig 9 (14.0), May 1978, LGA, B.E. Natural de Lima, Peru," 16* (28.6). (3) fUSNM, Felgenhauer, L. Deaton. Provincia de Panama—(10) Rio Chicama, near Trujillo, 16* (27.4), Jul 1970, fLGA, 12 km N of El Llano on Carti Rd, 66* F. Ancieta. (4) *USNM, purchased at market in (10.2-21.8), 29 (10.2, 12.7), 10 Mar 1973, Mc- Salaverry, 16* (26.7), 24 Oct 1926. (5) Rio Je- Phail, R.L. Dressier. (11) fUSNM, creek 1.7 km quetepeque, near Pacasmayo (Solar, 1972:8). (6) E of El Llano (Bravana Basin), 19 (16.0), 17 Mar Rio Chicama, near Pacasmayo (Solar, 1972:8). 1962, HL. (12) fLGA, Rio Mamoni (Bayano In addition to the rivers listed here, Mendez Basin), 26* (25.6, 26.1), 29 (10.2, 14.7), 16 Apr (1981:70) also reported its occurrence in rios 1973, LGA, RLD; 16* (29.0), 16 May 1973, LGA, Chira, Supe, and Huaura, but no specific locali- RLD; 26* (11.8, 13.1), Apr 1974, J. Gee, D. Kra- ties were cited. mer; 16* (20.0), 19 (13.3), 3 Aug 1974, LGA, K.L. There is every reason to believe that the follow- 98 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY ing locality data are erroneous: same is true of the two distal rows of spinules on ARGENTINA: *BM, Buenos Aires, Id* (38.0), 1$ the ultimate podomere of the peduncle. Varia- (13.4), March 1892, H. Pittier. (Perhaps pertinent tions in the arrangement of tubercles on the flexor is the fact that H. Pittier collected in Oaxaca, surface of the propodus and dactyl of the third Mexico; see locality 6 under "Mexico," above.) pereiopod are illustrated in Figures 37fji,i,k, 40b, NEW CALEDONIA: *MHNP, syntypes ofAtya mar- 41. The merus of the fourth and fifth pereiopods garitacea, 26 (24.5, 30.6), 1$ (16.6). of most small individuals bears two or three mov- AUSTRALIA: *MHNP, Victoria, 16* (21.8), von able ventral spines and a subterminal ventrolat- Miiller. eral one, and there is a large one present on or VARIATIONS.—The carapace of freshly molted near the distal ventrolateral extremity of the car- individuals is studded with a dense pile of short, pus; any one or all may be lacking in larger comparatively stiff setae, but frequently speci- representatives of the species. mens are encountered in which the pile is so None of the variations noted have been discov- abraded as to be hardly noticeable. Among other ered to be helpful in recognizing the geographical variations are features of the rostrum, the margins source of a specimen. Virtually all of the limits of of which may be very weakly or strongly concave, variations noted are exhibited by a series of spec- and whereas always angular at the base of the imens from a single Panamanian locality. acumen are sometimes produced with acute or ECOLOGICAL NOTES.—Ecological observations rounded apices. The apex of the acumen always pertaining to this shrimp are indeed few. Smith lies at or near the articulation of the basal and (1871) noted that his specimens of Atya rivalis penultimate podomeres of the antennule, and the were collected in freshwater streams, and Abele high to moderately low concave dorsal carina and Blum (1977) stated that it is found only in almost or quite reaches the apex; the ventral riffle areas of fast-flowing, clear streams. In their carina may lack or possess a small preapical tooth. study of the freshwater decapods in the Archipie- The posteroventral angle of the pleuron of the lago de las Perlas, the latter found that A. rivalis third abdominal segment may be angular or seemed to be limited to the larger islands where rounded, and the number of denticles on the such a habitat exists throughout the year. ventral margin of the third, fourth, and fifth Of the specimens examined by us, all of them pleura are quite variable, at least in part due to that are accompanied by data other than that loss of one or more of them. The number of spines citing the locality include a reference to a creek, in each of the curved rows on the dorsal side of river, or stream. The greatest altitude at which the telson varies from five to seven, and the specimens are known to have been obtained was posteromedian tubercle, always premarginal, in Panama where they were found at 426 m may or may not overreach the caudal margin. above sea level. The preanal spine appears usually to be propor- We have discovered no comb-bearing scrapers tionately longer in smaller individuals; with re- among the terminal clusters of setae on the fingers spect to the dorsal base of the carina, becoming of the first and second pereiopods of this shrimp, shorter with increase in size of the animal. The and the terminal brush on the "bristles," like basal podomere of the antennule may bear one, those of A. scabra, are comparatively inconspi- two, or no corneous spinules among the plumose cuous as compared with those of A. innocous. As setae on the dorsal surface, and the number of pointed out in the introductory section of this spinules in the transverse distal row varies from study, Fryer (1977:93) associated these features six to 10; even more variable is the number of with the feeding behavior of the shrimp, correlat- spinules on the dorsodistal margin of the penul- ing (p. 73) such bristles as those possessed by A. timate podomere; frequently gaps occur in both margaritacea with predominance of filtering over rows suggesting a loss due perhaps to injury. The scraping. NUMBER 364 99

FIGURE 40.—Atya margaritacea (all from male from Mt. Chiriqui, Panama): a, dorsal view of cephalic region; b, flexor surface of distal part of third pereiopod; c, lateral view of cephalic region; d, lateral view of second through fifth pleura; e, lateral view of preanal carina;/, lateral view of third pereiopod; g, dorsal view of telson; A, mesial view of appendices masculina and interna. (Scales marked in 1 mm increments.)

LIFE HISTORY NOTES.—Even less is known of 11.4 mm, 3010 eggs. Such a high biotic potential the life history of this shrimp than about the suggests that like Micratya poeyi and Atya innocous, habitats in which it lives. Ovigerous females have studied by Hunte (1979a,b), which require an been collected in April, May, July, September, increased salinity in their early larval stage, the October, and November. The numbers of eggs hatchlings of members of this species probably carried by three females were as follows: carapace travel downstream for at least a temporary so- length 9.3 mm, 1770 eggs; 9.5 mm, 1798 eggs; journ in an estuarine, if not marine, habitat. The 100 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 41.—Atya margantacea, variations in flexor surface of distal part of third pereiopod (numbers in parentheses * carapace length in mm): a, Rio Machiro, San Mateo, Mexico (6, 28.5); b, San Ramon, Mexico (9, 33.5); c, Pentalion, Guatemala (6", 22.7); d, Rio Riryto, Costa Rica (<$, 17.5); e, Rio San Martin Grande, Panama (c$, 25.6);/, Rio Chicamo, Peru (6, 27.4). NUMBER 364 101 dimensions of the eggs carried by the syntypic Arteaga, Michoacan, Mexico; types: IBM 21153, holotype female of A margaritacea were reported by Bouvier (9), allotype (6); paratypes, USNM 99527 (16\ 1$); Insti- (1925:315) to be 0.51 by 0.31 mm, measurements tuto Politecnico Nacional (2 specimens)]. that are in keeping with those noted among the The above citation is the only reference that ovigerous females examined by us. we have encountered to this shrimp; it consists of REMARKS.—Because of the confusion surround- a description and illustrations that were based ing the identity and range of this shrimp, we are upon four males and five females from the type- illustrating both a syntypic male of Millie-Ed- locality. Villalobos did not account for three of wards' species and a male specimen from the the specimens listed in his measurements but Pacific watershed of Guatemala. The syntypes of which were not included among the types (see Atya rivalis are in such poor condition, and we pp. 472, 474). have no specimens from Nicaragua, that this PUBLISHED ILLUSTRATIONS.—The excellent il- specimen from Guatemala has been chosen to lustrations provided with the original description represent Smith's species. Inasmuch as the syn- depict a lateral view of the entire female, a lateral types of A. margaritacea appear to us to be indistin- view of the carapace, dorsal view of the cephalic guishable from the syntypes of A. rivalis as well as region and of the telson and uropods, drawings from other specimens collected on the Pacific of the gnathal and ambulatory appendages, the versant from Baja California to Peru and are first and second pleopods of the male, and mesial obviously different from specimens of Atya scabra views of the appendix interna and appendix mas- from throughout its range, we are convinced that culina. Setae from the scaphognathite, second the name Atya rivalis should be removed from the maxilla, and first and second pereiopods, includ- synonomy of A. scabra and treated as a junior ing those with biserial arrangement of denticles, subjective synonym of Milne-Edwards' Atya mar- are also presented. garitacea. DIAGNOSIS.—Cephalic region of carapace not In our opinion, A. margaritacea is more closely conspicuously sculptured, with anteriorly con- related to A. scabra than to any other member of verging flat (not convex), glabrous lateral sur- the genus. It may be distinguished from the latter faces; antennal and pterygostomian spines prom- by several features, among the most conspicuous inent. Rostrum projecting anteriorly, not arched of which are (1) the arrangement of the tubercles or directed ventrally, with margins suddenly con- on the flexor surface of the propodus of the third tracted at base of acumen forming angular to pereiopod; in the former the tubercles are scat- subangular bends; angles never produced ante- tered over the entire surface, and if a few tend to riorly; acumen longer than remainder of rostrum. lie in a linear series they are never contiguous or Ventral margin of third through fifth abdominal overlapping (cf. Figures 41, 53), (2) the distome- pleura lacking rows of sclerotized denticles; cau- sial extremity of the merus of the third pereiopod doventral angle of fourth and fifth pleura weakly is produced in a prominent rounded lobe in A. obtuse to acute but not produced. Sternum of margaritacea, one that is far more conspicuous than sixth abdominal segment distinctly more than the comparatively small corresponding bulge in half as long as wide; compressed median tubercle A. scabra, and (3) spinules are never present on on sternum of fifth abdominal segment small and the ventral margin of the second abdominal pleu- comparatively inconspicuous. Preanal carina ron in A. margaritacea but in the Western Hemi- with compressed spine reaching to or beyond sphere are rarely absent in A. scabra. caudal extremity of basal part of sclerite. Telson Atya ortmannioides Villalobos slightly more or less than twice as long as wide with 6 to 8 denticles in each of 2 dorsal rows. FIGURES \b, 9, 10, \2c, 42-44 Antennular peduncle with dorsal surface of prox- Atya ortmannioides Villalobos, 1956:459-475, pis. 1-6 [type- imal article devoid of sclerotized denticles proxi- locality: Rio de las Truchas, La Mira, 52.5 km SSE of mal to distal row; penultimate article 1.5 to 1.8 102 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 42.—Atya ortmarmioides (all from paratypic female): a, dorsal view of telson; b, dorsal view of cephalic region; c, flexor surface of distal podomeres of third pereiopod; d, lateral view of cephalic region; e, preanal carina;/, lateral view of third pereiopod; g, lateral view of second through fifth pleura. (Scales marked in 1 mm increments.)

times as long as wide, and dorsal surface with MALE (RIO Murga, Estado de Guerrero, Mex- about 15 scattered spinules. Coxae of third and ico).—Rostrum with margins strongly contracted fourth pereiopods lacking prominent anterolat- at base of acumen forming angle of slightly more eral spines. Third pereiopod with merus rounded than 90 degrees; acumen about 1.5 times as long ventrally, approximately 5.5 to 6.5 times as long as basal part of rostrum and reaching distal end as high, entire podomere bowed, never parallel to of proximal third of penultimate podomere of merus of other member of pair, and lateral surface antennule; dorsal median carina gently curved, bearing weakly sclerotized (none cornified) tuber- not excavate dorsal ly (not dipping below level of cles, some of which sublinearly arranged; propo- lateral carinae posterior to acumen), and reaching dus about 3 times as long as wide, its extensor base of corneous tip of acumen; ventral carina surface studded with acute to subacute cornified with single small preapical tooth; ocellar beak denticles, and flexor surface with similar denti- largely hidden by eyes, falling short of midlength cles, some linearly arranged but those in rows of stylocerite, its cephalic margin vertical, and neither contiguous nor overlapping; flanking se- dorsal one subtruncate, abutting shallow ventral tae not conspicuous; dactyl freely movable and rostral groove. Antennal spine strong, pterygos- bearing 2 irregular rows of sclerotized denticles. tomian spine less conspicuous. Surface of cara- NUMBER 364 103

FIGURE 43.—Atya ortmannioides (all from male from Rio Murga, Estado de Guerrero, Mexico): a, prcanal carina, sternum of sixth abdominal segment, and median tubercle of fifth; b,c, flexor surface of distal part of fourth and fifth pereiopods, respectively; d, lateral view of cephalic region; e, preaxial view of distal part of fifth pereiopod;/, dorsal view of antennular peduncle; g, mesial view of appendices masculina and interna; A, lateral view of third pereiopod; ij, flexor surface of distal part of right and left third pereiopods. (Scales marked in 1 mm increments.) pace glabrous, densely but finely punctate, punc- lacking denticles on ventral margin and fifth with tations bearing very short fine setae except just moderately conspicuous fringe of plumose setae. dorsal to ventral marginal sulcus where setae Fourth abdominal tergum 1.2 times as long as long; ridges and spines other than those just fifth, latter subequal in length to sixth and almost mentioned lacking. 0.9 as long as telson. Sternum of fifth abdominal Pleura of first 4 abdominal segments with segment with small compressed median tubercle rounded to subacute posteroventral extremities, (Figure 43a); that of sixth 1.2 times as broad as corresponding part of fifth forming almost right long. Free part of preanal carina spiniform, angle, not produced. Third through fifth pleura curved, and only slightly overreaching angle of 104 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY basal part of sclerite. Telson about twice as long tral surfaces of merus studded with sublinear as broad, its dorsal surface bearing paired concave series of prominent but noncornified tubercles rows of 6 corneous denticles and posteromedian flanked by arcs of plumose setae; moderately tubercle, latter overreaching caudal margin of conspicuous longitudinal band of short setae pre- telson. sent laterally; mesial extremity of podomere pro- Proximal podomere of antennule (Figure 43/) duced in small rounded lobe at level of mesial with stylocerite reaching about distal fifth of articular condyle of carpus. Latter bearing strong segment; dorsal surface with linear cluster of setae tubercles capped with corneous discs, tubercles but lacking corneous denticles; distal margin flanked distally by arcs of plumose setae, setae on bearing row of 13 (right) or 12 (left) corneous ventrolateral surface somewhat longer but not denticles; penultimate podomere about 1.5 times forming conspicuous tufts. Propodus also strongly as long as wide and provided with 15 (right) or tuberculate, tubercles capped with corneous 16 (left) corneous denticles on dorsal surface and spines; some tubercles on flexor surface forming row of 18 on distal margin; ultimate podomere mesial and lateral series and with more irregularly with 3 (right) or 2 (left, one of which broken) dispersed ones between, those in rows neither denticles on dorsal surface and row of 11 (right) contiguous nor overlapping; lateral and ventro- or 9 (left) at base of lateral flagellum and 4 at lateral surfaces with tufts of plumose setae. Dactyl base of mesial flagellum. Antenna with ventrolat- movable, its flexorsurfac e with denticles forming eral spine on basis reaching much farther ante- 2 irregular oblique rows flanked by paired setal riorly than pterygostomian spine but not quite so clusters. far as stylocerite; lateral spine on scaphocerite Fourth pereiopod with dactyl reaching end of almost reaching lateral extremity of antennular proximal fourth of propodus of third pereiopod; peduncle; lamella far surpassing latter; flagella merus about twice as long as carpus and latter broken but in accompanying female reaching only slightly shorter than propodus. Fifth pereio- third abdominal tergum. pod reaching distal end of proximal fourth of Third maxilliped only slightly overreaching propodus of fourth pereiopod, merus 1.4 times as antennular peduncle; tip of exopod attaining base long as carpus and latter 0.65 length of propodus. of distal podomere of endopod; penultimate seg- Ornamentation of merus, carpus, and propodus ment about 1.1 times as long as ultimate. of fourth pereiopod consisting of distal ventrolat- First pereiopod reaching base of distal third of eral spine and 1 more proximal ventral spine on penultimate podomere of antennular peduncle, merus, and distal ventrolateral and 2 smaller second almost attaining end of same podomere. distolateral spines on carpus; lateral surface of Terminal brush of both fingers with some bristles distal part of merus and of carpus and propodus bearing scraping denticles. Third pereiopod (Fig- with longitudinal narrow band of plumose setae. ure 43h-j) with lateral distoventral spine on Ornamentation of corresponding podomeres of merus and carpus, ventral spine lacking on both fifth pereiopod similar but with 3 distolateral podomeres; when appendage extended anteriorly spines on carpus instead of 2 and lacking setal overreaching antennular peduncle by dactyl and band on distal part of merus. slightly more than half length of propodus. Merus Diaresis of lateral ramus of uropod flanked with ventromesial margin bowed, about 5.4 times proximally by row of 20 (right) or 19 (left) artic- as long as high, 2.3 times as long as carpus, and ulated, corneous denticles, and fixedspin e present 2.2 times as long as propodus. Latter 3.1 (right) at lateral end of row. or 3.3 (left) times as long as wide and 0.9 as long COLOR NOTES.—No observations on the color as carpus; distoventral margin of coxa entire, of this shrimp have been reported. mesial caudoventral prominence and distal ven- SIZE.—Among the measurements recorded by trolateral spine lacking. Lateral, dorsal, and ven- Villalobos (1956:472) of four males and five fe- NUMBER 364 105 males were the total length and carapace length pulco, 1$ (14.2), 22 Jan 1895, Nelson and Gold- (the latter was apparently made from the tip of man. (3) flBM, Rio Murga, 16 km de la carretera the rostrum to the posterior margin of the cara- Petatlan-Zihuatanejo, 36* (9.7-24.6), 1$ (15.0). G. pace). His largest specimen was a female having Malagrino J. Estado de Michoacdn—(4) *IBM, Rio a carapace length of 21.0 mm and a total length de las Truchas, La Mira (Villalobos, 1956:474), of 59.0 mm. Corresponding measurements of the holotypic $ (15.9), paratypic 6 (8.8); *USNM, largest male were 11.5 and 32.7 mm. Among the paratypic 9 (12.9), paratypic 6 (7.7), 11 Feb 1953, specimens measured by us, the largest was a male B.F. Osorio-Tafall, R. Ramirez, T. Gutierrez. having a carapace length (measured from the VARIATIONS.—The variations noted in the spec- orbit to the posterior margin of the carapace) of imens of this shrimp just listed are hardly note- 24.6 mm, and that of the largest female was the worthy. In all of them, the rostrum reaches to at holotype with a carapace length of 15.9 mm. No least the base of the penultimate podomere of the ovigerous females have been reported. antennule, and the acumen, although poorly de- DISTRIBUTION AND SPECIMENS EXAMINED.—This limited (the angles rather strongly obtuse) basally shrimp has been found in four localities between in the smaller specimens, is clearly longer than Estado de Guerrero, to La Paz, Baja California. the basal part. The spine on the ventral carina Records for the known localities are listed be- may or may not be present. As might be expected, low. Collections that we have examined are fewer denticles are present on the antennular marked with an asterisk if they have been previ- peduncle of the small individuals than on those ously reported and with a dagger if they are of the comparatively large male described. As has reported herein for the first time. Numbers follow- been observed in other species, the number of ing the specimens listed are measurements, in spines on the ventral surface of the merus of the mm, of the carapace length. One lot lacks the ambulatory appendages, especially that of the date it was collected; this could not be deter- fourth and fifth pereiopods, in small specimens is mined. greater than in larger individuals; in A. ortman- MEXICO: Estado de Baja California Stir—(1) nioides the small individuals have three, whereas fUSNM, La Paz, 26 (9.9-10.0), Jun 1882, L. the male that is described has only one. Belding. Estado de Guerrero—(2) fUSNM, Aca- We are unable to discover a single feature in the two specimens from Baja California Sur that serves to distinguish them from the specimens from the more southern localities in Guerrero and Michoacan. ECOLOGICAL NOTES.—Except that two of the lots examined by us bore labels noting that the specimens had been collected in rivers, there are no ecological data available other than those recorded by Viltalobos (1956:474). The following is a free translation of the information quoted from the collectors by Villalobos: The exact place where the shrimp were collected lies at the foot of Las Truchas hill, which seems to be the site of the most important iron-ore deposit in Mexico. The river before entering onto the alluvial plain courses through a series of canyons carved through granite, forming rapids, small cataracts, FIGURE 44.—Distribution of Atya ortmannioides. and extensive potholes. At the edge on which the 106 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY slope begins, in the rapids and cataracts, plants America, the form that was designated A. tenella are firmly rooted among the rocks, and their by Smith. As shown by Fryer (1977), A. innocous leaves are constantly drenched by the flowing also exhibits dentiform bristles among the termi- water. The crustacean (shrimp) clings firmly to nal setal clusters of the first and second pereio- the leaves of these plants. In order to dislodge it, pods. The males available to Villalobos were it is necessary to shake the leaves over a ring net rather small (carapace lengths, 7.7 and 8.8 mm, or, exerting pressure, to run the net over the mats tip of rostrum to posterior margin of carapace) of such plants several times. In this section of Las and perhaps juvenile; if so, it is not surprising Truchas River, the so-called trout, which are that the spines on the margin of the endopod of freshwater mugilids belonging to Agonostoma mon- the first pleopod were not evident. The spines are ticota, were also collected. Moreover several large present on the largest male examined by us. specimens of Macrobrachium were obtained. The most distinctive features of this shrimp Whereas there are records of other Atya occur- occur in the rostrum and in the dactyl of the fifth ring among aquatic vegetation, this is the only pereiopod. The rostrum extends almost horizon- account of their clinging to leaves of plants. tally instead of being arched or directed antero- Whether or not a search for specimens among the ventrally, and the acumen is longer than the rocks was undertaken was not stated, thus one basal part. The consistent pectinate arrangement cannot be certain that this microhabitat of Atya of the denticles on the flexor surface of the dactyl ortmannioides is occupied to the exclusion of others of the fifth pereiopod seems to be unique among by this shrimp. the American members of the genus, but such an LIFE HISTORY NOTES.—No ovigerous females arrangement has been observed on a few individ- have been reported, and nothing is known about uals of A. innocous and on several specimens of aspects of its life cycle. other species. Slightly less conspicuous are the REMARKS.—Villalobos (1956:474) compared anteriorly converging, rather than arched, lateral Atya ortmannioides with A. scabra and A. occidentalssurfaces of the cephalic region of the carapace of (= A. innocous), stating that it differs from them A. ortmannioides. As for the apparently unique in (1) lacking any sculpture of the carapace and features of the appendix masculina, there are tod the proportions of the latter, (2) by the strong few males available to be confident that such development of the abdominal pleural region, (3) features are unique. by the sharp, long rostrum, (4) by the presence of dentiform bristles among the terminal clusters of Atya scabra (Leach) the first and second pereiopods, (5) by the pecti- nate series of spines on the dactyl of the fifth FIGURES lk,l, 8-10, \lb,c, 45-53 pereiopods, (6) by the absence of spines on the Guaricuru. — Marcgrave, 1648:187, unnumbered fig.; external margin of the endopodite of the first 1942:187, unnumbered fig.—Jonstonus, 1650:35, 36, pi. pleopod of the male, and (7) by the form of the ix: fig. 16; 1657:27, 28, pi. ix: fig. 16; 1665:27, 28, pi. ix: appendices masculina and interna of the second fig. 16; 1767:35, 36, pi. ix: fig. 16.—Piso, 1658:78.—Jon- pleopod of the male. ston, 1660, pi. ix: fig. 16.—Ruysch, 1718:27, 28, pi. ix: fig. 16.—Sawaya, 1942:LXIII.—Lemos de Castro, 1962:41, pi. Of these features, except for certain propor- 4: fig.26 . tions, the first is not unique, and many specimens Poti — Piso, 1658:78, fig; 1957:188, fig. of A. innocous lack any visible sculpture other than Gaurikuru.—Jonston, 1660:27. punctations that are also present in A. ortman- Guarikuru.—Jonston, 1660:27. nioides; the pleural region seems to be no more Atys scaber Leach, 1815:345 [type-locality: unknown ("Hab- strongly developed than are those of some A. itat "; Leach, 1815:345); restricted to "the area of Veracruz, Mexico" (Holthuis, 1966:234); fixed by the innocous, particularly those of members of the present neotype selection to Misantla, Estado de Veracruz, species occurring on the Pacific versant of Middle Mexico, 19°56'N, 96°50'W; types: not known to exist, NUMBER 364 107

ncotype designated herein].—Holthuis, 1955a:208, 221; Davant, 1963:42-44,98-100, figs. 27,28 [both figs, on pp. 1955b:26; 1957:167, 180.—Hemming, 1957:133, 150, 167, 43 and 49].—Chapa, 1964:34.—Pericchi Lopez, 1965: 180. 25.—Holthuis and Rosa, 1965:9.—Pinchon, 1967:161.— Atya scabra.—Leach, 1816, pi. 21; 1817:29, 30, pi. 131; Coelho and Ramos, 1968:3.—Fausto Filho, 1968:28.— 1824:421, pi. 21.—H. Milne Edwards, 1816:138, 139, pi. Straskraba, 1969:17, 25.—Chace and Hobbs, 1969:5, 51: fig. 4; 1837:348, pi. 24: figs. 15-19; 1838:352; 15, 19, 33, 36, 57, 61, 63-66 [in part], 73, figs. 9, lOd-f, 1839:386; 1840, pi. 51: fig. 4-4e; 1842:138.—Latreille, 14d,e— Cendrero, 1971:524.—Disney, 1971:84,85; 1975: 1817:37.—Desmarest, 1823:313, pi. 37: fig. 2; 1825:217, 69.—Hobbs, 1971:27.—Chace, 1972:14.—Lemasson, pi. 37: fig. 2.—Guilding, 1825:338.—Guerin-Meneville, 1973:68.—Bonnelly de Calventi et al., 1973:1338; 1974: 1825:329; 1837, vol. 2, pi. 20: fig. 2; 1844, vol. 3, p. 15; 104, 106, 111, 123, fig. 17.—Leveque, 1974:42, fig. lj.— 1856-I857a:xlvi, xlvii, pi. 2: fig. 7c; 1856-1857b:xviii, pi. Bonnelly de Calventi, 1974b:35, 38, 40, 54, fig. 7.— 2: fig. 7c—P. Roux, 1831:27.—Lucas, 1833-1834:333; Alayo, 1974:22, pi. vi: fig. 11.—Scelzo, 1974:12.—Abele, 1840:182, pi. 11: fig. 1; 1842:182, pi. 11: fig. 1; 1851:182, 1975:56, 57.—Hume, 1975:66; 1978:135, 136, 139, 144, pi. 11: fig. 1.—Griffith and Pidgeon, 1833:190.—Broderip, 145,147, fig. 7; 1979c: 70.—Carvacho and Carvacho, 1976: 1835:71, fig.—Griffith, 1835:223, lxiii.—Wiegmann, 213, pi. 2: fig. 8.—ViUamil and Clements, 1976:5, 59.— 1836:146, 147.—White, 1847:74.—Newport, 1847:158, Fryer, 1977:57, 58, 62, 63, 72, 73, 90,93, 94, 98, 125, figs. 159.—Stimpson, 1857:498.—A. Milne-Edwards, 1864: 3, 45, 49, 50, 52, 74, 79-82.—Burkenroad, 1981: 261.— 146, 147.—Costa, 1864:89.—Capello, 1867, explanation Rodriguez, 1981:46. of pi. 1: fig. If.—Giebel, 1875:53, 54.—Brocchi, 1875:32, Atya Scabra.—Leach, 1816:421.—A. Milne-Edwards, 1864: figs. 33, 34.—Vilanova y Piera, 1875:385.—Kingsley, 150.—Valdes Ragues, 1909:180.—Hart and Hart, 1878a:92; 1878b:56, 57.—Greeff, 1882:35-37; 1884: 1974:142. 54.—Albert, 1883a:469-471, 534, pi. 29: figs. 4-6; Alia scabra.—Latreille, 1817:37 [erroneous spelling, combi- 1883b:27-29.—Mocquard, 1883:190-199, 307, pi. 8: figs. nation by implication]; 1837:xvii, Crustacea pi. 31 (bis): 183-191.—Claus, 1885:57, 107, pi. 4: fig. 35.—Gundlach, fig. 2 [erroneous spelling]. 1887:131.—Osorio, 1887:222, 230; 1888:188; 1889:129, Atie epineuse.—Audouin and H. Milne Edwards, 1829:14, 137, 139; 1891a:47; 1891b: 140; 1892:200; 1895a:249; pi. 25: fig. 1. 1895b:251; 1898:186, 187, 194; 1905:102; 1906:150.— Atya.—Broderip, 1835:71.—Capello, 1866:6.—Gerstaecker Bate, 1888:693, 694, 698.—Pocock, 1889:16.—Ortmann, and Ortmann, 1881-1901:886, 887, 1127.—Johnston, 1890:466; 1895:408-410, 415,416 [in part]; 1897:183-185 1906:843.—Sawaya, 1942:LXIH.—Darnell, 1962:440. [in part].—Sharp, 1893:111.—Stebbing, 1893:240.—No- Astacus (Atya) scabra.— Voigt, 1836:178. bili, 1897:5.—Rathbun, 1897:44; 1900:313-314 [in part]; Atya mexicana Wiegmann, 1836:145 [type-locality: Misantla, 1901:119 [in part].—Dofiein, 1900:127.—Young, 1900: Estado de Veracruz, Mexico; types: ZBM, 2 specimens].— 473.—Thompson, 1901:22.—Bouvier and Lesne, 1901:13 Newport, 1847:158, 159.—Guerin-Meneville, 1856-1857a: (or 332).—Gerstaecker and Ortmann, 1881-1901, pi. 91: xlvi.—A. Milne-Edwards, 1864:147.—Martens, 1868: figs. 5, 6, 6a; pi. 115; figs. 8-11.—Bouvier, 1904:138 [in 49.—Giebel, 1875:52.—Bate, 1888:693.—Holthuis, 1966: part]; 1905:110, 112, 119-123, 128 [in part], fig. 25s; 234. 1906:492; 1909:333; 1925:22, 27, 29, 293, 312-319, Athys scabra.—H. Milne Edwards, 1838:352; 1839:386 [erro- 322, 323, 356, 358 [in part], figs. 55-67.—Johnston, 1906: neous spelling]. 862.—Balss, 1914:97, 98; 1925:239.—Tesch, 1914: "Atya sulcatipes?" Newport, 1847:159, pi. 8: fig. 1 [type- 247, 250.—Pearse, 1915:551.—Torralbas, 1917, fig. 60.— locality: Sao Nicolau, Cape Verde Islands; types: stated De Man, 1925:27-28 [in part], fig. 4a-c.—J. Roux, to be in the British Museum but could not be located in 1926a:238, 253; 1926b:217, 218.—Allee and Torvik, April 1980]. 1927:67—Monod, 1928:121, 205; 1933:461,462 [in part]; Atya sulcatipes.—White, 1847:74.—A. Milne-Edwards, 1967:110, 119, 135, pi. ix: figs. 11, 12, 15-17 [not figs. 13, 1864:147.—Martens, 1868:49.—Giebel, 1875:52, 55.— 14 as noted in explanation of plate]; 1980:375, 376, figs. Gerstaecker and Ortmann, 1881-1901, pi. 73: fig. 6a-c; 5, 6.—Schmitt, 1935:135, 136 [in part].—Villalobos, pi. 74: figs. 4, 5.—Bate, 1888:693-699, 701-704, pis. 118, 1943:1-67, pis. 1-22; 1956:474; 1959:328, 329.—Oliveira, 119: fig. 1, lp, lq.—Hickson, 1889:222, 223, 362.—Steb- 1945:177-189 [in part], pi. 1: figs. 1-4; pi. 2.—Holthuis, bing, 1893:240.—Oliveira, 1945:179,180,188.—Holthuis, 1951:9, 22, 24, 25 [in part]; 1955b:26; 1959:16; 1966:233, 1966:232-237, fig. 4; 1980:71, 181.—Rutherford, 1971:87, 234, 238 [in part]; 1974:231; 1980:70, 181.—Parodiz, 88, 90, fig. 2a,b.—Disney, 1975:69. 1960:38, 39 [in part], figs. 1-3.—Rioja, Ruiz, and Larios, Atya Mexicana.—Stimpson, 1857:498. 1955, fig. 316.—Darnell, 1956:131-138, 3 figs.; 1962: Atya scalva.—Martens, 1872:135 [erroneous spelling]. 440.—Hart, 1961b:61-64, 67, 73, fig. 11; 1980:845, Atya scaber.—Giebel, 1875:52 [erroneous spelling]. 846, 848.—Lemos de Castro, 1962:41, pi. 4: fig. 27.— Atya punctate Kingsley, 1878a:91-92 [type-locality: Haiti; 108 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

type: ?USNM 84327, $]; 1878b:57.—Pocock, 1889:16.— to Herbst's "cane, innocous." Desmarest (1823 Oliveira, 1945:179.—Holthuis, 1966:237.—Chace and and 1825) added no new information but pre- Hobbs, 1969:63. sented a lateral view of an entire animal in both Atya sculptipes.—Ortmann, 1890:466 [erroneous spelling]. publications. The earliest locality cited is that by Atya margaritacea var. claviger Aurivillius, 1898:14-16, pi. 3: figs. 5-8 [type-locality: "Etome, in Bachen," Cameroon; Guilding (1825) who recorded the presence of the types: SMNH, 2 specimens]. species in mountain streams of Saint Vincent "in "Atya scabra, type de margaritacea."—Bouvier, 1925, figs. 703, incredible numbers." Whereas there is no reason 704 [not 705, 706 as indicated]. to doubt that the species occurs on this island, we Atya sacabra.—Villalobos, 1943:66 [erroneous spelling]. suspect that most of the shrimps that Guilding Atya margaritaria clavipes.—Holthuis, 1966:234 [erroneous spelling]. identified as Atya scabra were members of the more Atyia scabra.—Velez, 1967:42.—Bonnelly de Calventi, common A. innocous. Guerin-Meneville (1825) pre- 1974a: 16.—Alayo, 1974:25 [erroneous spelling]. sented a brief diagnosis of his "Atye Raboteuse," Atya innocous.—Chace and Hobbs, 1969:66 [lapsus for Atya and in a subsequent work (1829-1844) included scabra]. a lateral view of a specimen and reported (p. 15) Atya sulzatipes.—Rutherford, 1971:90 [erroneous spelling]. the presence of the species in "les cotes du Me- Atya sukaltipes.—Rutherford, 1971:89 [erroneous spelling]. Atya scubra.—Bonnelly de Calventi, 1974b:54 [erroneous xique." Polydore Roux (1831), Griffith and Pid- spelling]. geon (1833), Lucas (1833-1834), Broderip (1835), Griffith (1835), and Voigt (1836) added nothing REVIEW OF LITERATURE.—The first record of to our knowledge of the species. Wiegmann (1836) the existence of a member of the genus Atya is described this shrimp, employing the name Atya that of Marcgrave (1648), who recorded it by the mexicana, from Misantla, Veracruz, Mexico. The vernacular name "Guaricuru" in his "Historia following year, H. Milne Edwards (1837) cited Naturalis Brasiliae," presenting an illustration of previous references, briefly described the species, an unmistakable member of the genus. Lemos de and under "Habite" listed only "cotes du Me- Castro (1962) pointed out that this figure almost xique." The latter, together with the similar ref- certainly depicts a representative of Atya scabra. erence by Guerin-Meneville, led several subse- The description, which is in Latin, was repub- quent authors to assume that this shrimp fre- lished by Jonstonus (1665) along with a slightly quents marine habitats. Nothing new related to modified figure. A. scabra appeared in H. Milne Edwards' 1838 This shrimp, from an unknown locality, was and 1839 publications, but in that of 1840 he briefly described by Leach (1815) as Atys scaber; included a hand-colored lateral view of a speci- the whereabouts of the type, if extant, is also men along with detailed illustrations of several unknown. Apparently Leach discovered that the body parts. Latreille (1837) also presented a name "Atys" had been employed by De Montfort hand-colored dorsolateral view of an entire ani- (1810:343) for a gastropod mollusk, and in 1816 mal. Lucas (1840, 1842, 1851) included addi- Leach illustrated and referred to his shrimp as tional features in his diagnosis of the species and Atya scabra. The latter designation was adopted presented a lateral view of an entire animal; the same year by H. Milne Edwards (1816:138), otherwise he added nothing new. Newport (1847), who presented an expanded description and well- apparently with hesitation, described Atya sulca- executed illustration of a specimen. The only tipes from San Nicolas, Cape Verde Islands, stat- addition of significance offered by Leach (1817) ing that "it may prove to be only a variety" of is the hand-colored dorsosinistral view of an entire Atya scabra. In the same year, White (1847), in animal. Latreille (1817) introduced the synonym listing the Crustacea in the British Museum, in- Atia in stating that although Leach's species was cluded A. scabra from Mexico (from the Leach unknown to him, it "appears to connect the collection) and another lot from the Cape Verde penaeids and alpheids." He also called attention Islands (from the collection of Newport). Whereas NUMBER 364 109 he listed A. sulcatipes, he gave no locality and bert (1883a) also provided an anatomical descrip- made no comment concerning the collection. tion of this organ. In an excellently illustrated Guerin-Meneville (1856-1857a,b) contrasted this study of the branchial apparatus of crustaceans, shrimp with his Atya poeyi (= Micratya poeyi). Claus (1885) presented the gill formula of A. Stimpson (1857) "ventured to cite Wiegmann's scabra and a well-executed drawing of the bran- name [Atya mexicana] as a synonym" of A. scabra. chiae. The presence of this shrimp in a tributary To Wiegmann's locality, he added western Mex- of Rio Anasco and in the San Juan, Puerto Rico, ico as a new record for the species, and he also market was noted by Gundlach (1887) who also expressed the belief that no member of the genus mentioned its occurrence in Cuba and Mexico. is found in the sea. On what basis the "western In a series of reports dealing with the fauna of Mexico" record is founded is not known; perhaps the then Portuguese West Africa (1887-1906), he had specimens of A. margaritacea or A. ortman-Osorio listed A. scabra from a number of localities nioides that were misidentified. That A. mexicana in Angola, Annobon, Ihla do Principe, and Sao and A. scabra are conspecific was pointed out by Tome. Bate (1888) reported A. sulcatipes from San A. Milne-Edwards (1864), who also stated that Antonia, "San Iago," Cape Verde Islands, pre- except for the occurrence of A. sulcatipes in the sented a detailed description accompanied by Cape Verde Islands, "one would not hesitate to illustrations, and stated that it differs from A. see in it a variety of A. scabra; in effect it seems to scabra "in having a prominent tooth on the inner differ only in the legs which are slightly grooved, distal angle of the carpus," (p. 694) and a longer by the rostrum which exhibits a more marked second antenna (p. 698). Hickson (1889) added depression, and by the antennae which are not no new information on the species. Pocock (1889) more than half the length of the body" (p. 147, commented that the only features selected by translation). In this work he presented a key to Kingsley (1878a) to distinguish his A. punctata all of the species belonging to the genus. Costa from other species were probably individual vari- (1864) and Capello (1867) provided no new in- ations or due to differences in age. formation. Von Martens (1868) added more com- Ortmann (1890) included Atya scabra and A. plete data for the specimens on which Wiegmann sulcatipes in his margaritacea group on the basis of described Atya mexicana: a small coastal stream the shape of the rostrum. In his study of the flowing into the Gulf of Mexico south of Tecolu- family Atyidae (1895), he grouped the following tla, about 20 km from the sea. Giebel (1875) in his synonomy of A. scabra'. A. mexicana Wieg- summarized the composition of the genus Atya mann, A. sulcatipes Newport, A. occidentalis New- and in his description of A. gabonensis made com- port, A. rivalis Smith, A. tenella Smith, and A. parisons with A. scabra. In Brocchi's study (1875) punctata Kingsley. This synonomy set the stage for of the genital organs of decapod Crustacea, he repeated erroneous interpretations of the species mentioned and figured the first and another (no involved. The range cited for A. scabra by him statement as to which) pleopod of A. scabra. was based, at least in part, as follows: Nicaragua Kingsley (1878a) described Atya punctata from on records for A. rivalis (— A. margaritacea) and A. Haiti and in the same year (1878b) listed A. scabra tenella (= A. innocous) by Smith (1871); Jamaica among the North American caridean shrimps, on that for A. occidentalis (= A. innocous) by New- reporting it from fresh water in western Mexico port (1847); and Dominica on that for A. occiden- (perhaps a misidentification of A. margaritacea); A. talis by Pocock (1889). The Haitian record is that mexicana was cited as a synonym. Greeff (1882, of Kingsley (1878a) for A. punctata, and the Cape 1884) reported the occurrence of A. scabra on Sao Verde Islands localities are those for A. sulcatipes Tome. Mocquard (1883), in an investigation of reported by Newport (1847) and Bate (1888). the stomachs of a number of crustaceans, de- Inasmuch as both A. scabra and A. occidentalis {— scribed and illustrated that of A. scabra, and Al- A. innocous) occur on Tobago, which of the two no SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY was examined by him is not known. Stebbing three species under A. scabra. Her statement that (1893) expressed the opinion that A. sulcatipes "is "this species is extremely variable in the form of perhaps only a variety of the older Atya scabra." the rostrum and more noticeably in the shape of Ortmann continued to recognize A. margaritacea the last three pairs of thoracic feet which may all as a distinct species in his 1895 report but stated: be slender and nearly equal in length; and their "The differences between the New Caledonian spines very feeble and appressed" suggests that species A. margaritacea and robust a and the West among her material were A. scabra, A. lanipes, and, Indian A. scabra are very doubtful..." (p. 408). in all probability, A. innocous. Specimens of A. Except for a diagnosis of the species in Portu- lanipes from Puerto Rico collected by the Fish guese, nothing new relative to A. scabra was re- Hawk in 1899 may not have been examined by corded by Ortmann (1897). The occurrence of Rathbun until much later, for they were not this shrimp in Rio Macuto, near La Guaira, entered in the Smithsonian catalogue until 1913, Venezuela, was noted by Nobili (1897). In the at that time identified by her as A. occidentalis. same year, Rathbun (1897) cited Jamaica as a Her summary of the distribution of A. scabra was locality for it, but her statement, "This species is almost identical to that recorded by her in 1900. probably identical with A. occidentalis Newport," The first report of the occurrence of Atya scabra leaves some doubt as to the identity of the mate- in Panama was that of Doflein (1900), who re- rial before her. Aurivillius (1898) described and corded it from the Atlantic watershed. No origi- illustrated A. margaritacea var. claviger, another nal data were offered by Young (1900), Thomp- synonym of A. scabra, from Cameroon. son (1901), or Bouvier and Lesne (1901). Ger- In her study of the West African decapods, staecker and Ortmann (1901) illustrated a lateral Rathbun (1900) erred in several respects: she view of an animal together with several append- apparently agreed with Ortmann's synonomy for ages. Bouvier (1904), considering A. margaritacea A. scabra adding A. gabonensis and Aurivillius' A. and A. sulcatipes as synonyms of A. scabra, recorded margaritacea var. claviger to the list. One of the the species from the following localities, several specimens cited from St. Paul River, Mount Cof- of which were new: New Caledonia; Victoria, fee, and that from Beulah were misidentified. The Australia; Darien, Colombia; San Esteban and Costa Rican record is new and may or may not Naricual, Venezuela; Valley Nacional, Oaxaca, be based upon a correct identification. All of the Mexico; Cape Verde Islands; Fernando Poo; and specimens from Costa Rica now in the Smithson- Sao Tome. In his synoptic treatment of this ian collection that would have been available to shrimp (1905), Bouvier included the following as her are members of A. margaritacea. The only synonyms: A. sulcatipes, A. rivalis, A. tenella, A. record of Atya from the Orinoco of which we are punctata, and A. margaritacea. This list differed from aware is that of Koelbel (1884) for Atya sculptilis Ortmann's in the recognition of A. occidentalis as (= A. gabonensis). Therefore, in all probability her a valid species and relegating A. margaritacea to inclusion of Costa Rica, Nicaragua, and Orinoco the synonomy of A. scabra. Its range, essentially in the range was based on erroneous determina- the same as that presented in 1904, differed only tions. The Liberian and "Gabun" records seem by the addition of the Antilles and west Africa. to have been founded, at least in part, on collec- He noted that many of the specimens from the tions of A. qfricana and A. gabonensis, respectively. Antilles probably belong to A. occidentalis and that The inclusion of Cameroon, Ihla do Principe, Sao several of the African specimens should be re- Tome, and Rolas, cited originally by other au- ferred to A. intermedia and A. qfricana. Atya margar- thors (and some based on records previously given itacea var. claviger was considered by him to be an for A. sulcatipes) have been confirmed. In her unquestioned synonym of A. scabra (p. 122). study of the Brachyura and Macrura of Puerto Nothing of importance was added to our Rico, Rathbun (1901) almost certainly treated knowledge of this shrimp by Bouvier or by John- NUMBER 364 111 ston in 1906, or by the former or Valdes Ragues ger are the only synonyms listed for A. scabra other in 1909. Balss (1914), following Rathbun, in- than Leach's original combination Atys scaber. No cluded in the range of the species the Atlantic new data were offered by Holthuis (1955b) in his versant of the Americas, the West Indies, and in synopsis of the recent caridean and stenopodidean and off Africa: Gabon, Sao Tome, Principe, An- genera; however, he (1955a) requested the Inter- nobon, and Rolas. Tesch (1914) predicted that A. national Commission on Zoological Nomencla- scabra would be found in Surinam. Pearse (1915) ture to place "Atys" of which Atys scaber Leach is recorded 79 specimens from Santa Marta, Colom- the type, on the "Official List of Generic Names." bia. Other than presenting a lateral view of the This was done by the Commission in Opinion entire animal, Torralbas (1917) offered no infor- 470 (Hemming, 1957). Villalobos' (1943) illustra- mation. tion of a specimen in lateral view was reproduced Bouvier's (1925) monograph of the family Atyi- by Rioja, Ruiz, and Larios (1955). In his descrip- dae has served as the principal source for all tion of A. ortmannioides, Villalobos (1956) con- subsequent work on members of the family. In it trasted it with A. scabra. Darnell (1956) analyzed he reviewed the previously reported localities, a population of A. scabra frequenting a small adding a number of new ones, and in his key and stream in Tamaulipas, Mexico, the most northern text contrasted and compared A. scabra with re- locality known for the species. In this study, he lated species. Balss (1925) reported the presence presented observations of their habitat, distribu- of the species in Victoria, Cameroon, and cited tion, population structure, food, and life history Valdivia as a new locality, and De Man (1925) and related his findings to those of Villalobos recorded its occurrence in the Belgian Congo. (1943). Jean Roux (1926a) stated that if the species occurs Holthuis (1959:16) pointed out that "Tesch's in Australia it must be rare. Working on Isla supposition that Atya scabra (Leach) might be Barro Colorado, Panama, Allee and Torvik found in Surinam has not yet been substanti- (1927) found this shrimp in Shannon Creek, and ated." Villalobos (1959) reported capturing ju- Monod (1928), in a report on the fisheries of venile forms of Atya scabra in the Rio Papaloapan, Cameroon, noted that A. scabra was of economic and stated that the earlier part of their develop- importance and pointed out its occurrence in the ment occurs near the coast, and afterward they Bimbia River. Later (1933), he recorded two new migrate upriver into areas far removed from the localities in Cameroon and cited a more precise mouths of the streams, in this respect resembling station on the Bimbia River. Schmitt's review Potimirim mexicana (De Saussure, 1857:505). Par- (1935) of the Macrura and Anomura of Puerto odiz (1960) cited a new locality for A. scabra in Rico contains no new data on the species. In a Brazil and provided illustrations. In his study of note on Marcgrave's treatment of "Guaricuru," the freshwater shrimps of Jamaica, Hart (1961b) Sawaya (1942) suggested that the shrimp was a presented photographs of a specimen and re- member of the genus Atya. ported three new localities. Darnell (1962) tabu- The beautifully illustrated morphological lated data from his previous study (1956). Atya study by Villalobos (1943), together with locality scabra was the only member of the family Atyidae and ecological data, is an invaluable contribu- considered by Davant (1963) to be of economic tion. Oliveira (1945) compared A. sulcatipes with importance in Venezuela, and he reported it to A. scabra, including a list of differences that sub- be a common inhabitant of the Rio Manzanares. sequently have not proven to be consistent. Hol- Lemos de Castro (1962), in reviewing the Crus- thuis (1951) presented the most nearly complete tacea included by Marcgrave in his "Historia synonomy previously available and cited all of Naturalis Brasiliae," called attention to the inclu- the localities offered by earlier investigators. In sion of "Guaricuru" with which Lemos de Castro this study, Atya sulcatipes and A. margaritacea clavi- associated Atya scabra. The illustration of Marc- 112 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY grave is reproduced and is accompanied by a also pointed out a number of characteristics in photograph of A. scabra from Rio Serinhaem, which this shrimp differs from the sympatric A. Estado de Pernambuco, Brazil. Holthuis and innocous. Disney (1971) noted that he had found Rosa (1965) noted that this shrimp is used for this shrimp in limited numbers in two localities food in Mexico, Costa Rica, Panama, Nicaragua, in the vicinity of Kumba, Cameroon: the Bille British Caribbean Federation, Cuba, Haiti, River (a tributary of the Meme River) and the Puerto Rico, Netherlands West Indies, Venezu- Blackwater River (a tributary of the Mungo ela, and Peru. The lattermost is based on the River). It was infested with the larvae and pupae misidentification of A. margaritacea. No new infor- of the blackfly Simulium atyophilum. Rutherford mation was added by Pericchi Lopez (1965). (1971) reported the occurrence of A. sulcatipes in Even though Holthuis (1966) found that the Ghana, noting that in addition to frequenting the differences between A. sulcatipes and A. scabra Cape Verde Islands, Sao Tome and Annobon, it compiled by Oliveira were not reliable when ranges on the African continent from Liberia to larger series were compared, on the basis of dif- Congo-Kinshasa. He stated that this dark green- ferences noted in the shape of the antennule, he ish gray shrimp occurs in rocky stream rapids. assigned the West African collections that had Cendrero (1971) added no new data. In consid- been designated as A. scabra by him in 1951, and ering the possibility that individuals of A. scabra earlier by Bouvier and others, to A. sulcatipes. He occurring in a small stream at Tapalapan, San- also considered A. margaritaria clavipes [sic] to be a tiago Tuxtla, Veracruz, serve as a host to the synonym of A. sulcatipes, and cited the known entocytherid ostracod Uncinocythere zaruri, Hobbs West African localities, including a number of (1971) reported a new locality for the shrimp. new ones, for the latter. The type of Atya scabra Lemasson (1973) pointed out that A. scabra attains was discussed by him, and he concluded that it is lengths of eight to 10 centimeters. Bonnelly de not among the specimens in the British Museum. Calventi et al. (1973) cited its presence in the The type-locality was restricted to "the area of Haina, Mana, Isa, and Nigua rivers and Canada Veracruz," Mexico. Atya mexicana, A. margaritacea, Madrigal in the Dominican Republic. They also and A. punctata were recognized as synonyms of noted that it "yields relatively little edible mass A. scabra. He also listed four differences between (13% total body weight) as compared with pe- the latter and A. rivalis (p. 234). Monod (1967) neids, but has a comparable content of water recorded A. scabra from the Cape Verde Islands, (74.5%) and protein (16.99%). Lipids represent Fernando Poo, Sao Tome, and Annobon; his 0.82-2.94%, ash 31.0%, calcium 0.232%, phospho- illustrations were taken from Bouvier (1925) and rus 1.033%, and iron 0.25%. Atya scabra muscle De Man (1925). Neither Velez (1967) nor Stras- provides 88.3 dietary calories per 100 g while kraba (1969) added to our knowledge of the peneid muscle provides 78.3." In their joint study species in pointing out its occurrence on Puerto (1974), Bonnelly de Calventi et al. pointed out Rico or Cuba, respectively. Except for recording that in the Dominican Republic this shrimp is the common name "Bouc," no new data were eaten only in some regions of the southern part of added by Pinchon (1967). Coelho and Ramos the country. Bonnelly de Calventi (1974a) listed (1968) reported that this shrimp is sold in the two additional rivers (Rio Jura and Rio Nizao) markets of Recife, Brazil, and Fausto Filho (1968) where it occurs in the Dominican Republic. In noted its use as a food item in northeastern Brazil. another report (1974b) she presented a diagnosis Chace and Hobbs (1969) followed Holthuis of the species, illustrations, notes on color, size, (1966) in their synonomy of A. scabra. They di- habitat, life history, and distribution in the Do- agnosed and illustrated the species, described the minican Republic. color pattern, reviewed its range, and commented Holthuis (1974) recorded a new locality in the on its ecological distribution on Dominica. They province of Pinar del Rio, Cuba. The following NUMBER 364 113

references to A. scabra contain no new data: Chace two centuries elapsed before Leach (1816) pre- (1972), Hart and Hart (1974), Scelzo (1974), sented a dorsodextral view of the shrimp. Much Hunte (1975), and Abele and Blum (1977). Le- better executed and more detailed illustrations veque (1974) noted the occurrence of this shrimp were furnished by H. Milne Edwards (1816); on the island of Guadeloupe and stated that these consist of a lateral view of the entire animal, ovigerous females had been found in June. It was a dorsal view of the cephalic region, the mandible, reported to be common in the eastern part of the first pleopod together with another unidenti- Cuba by Alayo (1974). Abele (1975) recorded the fied one, and a dorsal view of the sixth abdominal presence of the species in streams of the Atlantic segment, telson, and uropods. The following year, Basin in Panama, and, the following year, Vil- Leach (1817) published a hand-colored dorsosin- lamil and Clements (1976) stated that on Puerto istral view of a shrimp that appears to have been Rico it appears to be ecologically the most re- redrawn in reverse from the figure in his 1816 stricted and the most specialized of the crusta- publication. In his "Dictionary of Natural Sci- ceans occurring in the area. Fryer (1977), in his ence," Desmarest (1823) included a dorsodextral impressive study of the Dominican atyids, related view of an entire animal and in 1825 republished the presence on the chelae of the first and second the same figure. Guerin-Meneville's (1829-1844) pereiopods of bristles with scraping denticles to illustration of an entire animal was apparently "scraping" and "sweeping" for food. Their ab- redrawn from Desmarest's figure, Broderip's sence in A. scabra is correlated with the infre- (1835) from Leach (1817), and Latreille (1837) quency of its resorting to scraping; it usually used a hand-colored reproduction of Desmarest's employs filter techniques. He included a brief but illustration. The figures included in H. Milne exhaustive summary of available knowledge of Edwards (1837 and 1842) are the same as those their ecological distribution and habits. The ex- published by him in 1816. Lucas' (1851) illustra- cellent illustrations, particularly photographs of tion of an entire animal was apparently redrawn sections through the mouth parts and of a living from Desmarest (1823). A small likeness of the animal, greatly enhance his contribution. Among cephalic region in lateral aspect was depicted the most recent references to this shrimp are those by Guerin-Meneville (1856-I857a,b). Capello of Hunte (1978, 1979c). Although the latter in- (1867) included sketches of the first (or second?) troduces no new information, the former provides pereiopod. Mocquard (1883) and Albert (1883 not only additional locality records but also con- a,b) presented several figures illustrating the anat- tains valuable ecological data. Atya scabra was omy of the stomach of this shrimp. found by him on Jamaica only in the eastern Claus (1885) presented an excellent rendition parishes where there is a mountainous terrain. of the gills and basal parts of the thoracic ap- Data on tolerance of individuals to lowered tem- pendages. Bate's (1888) figures of A. sulcatipes perature and oxygen concentrations were offered include a lateral view of an entire animal, others in his effort to determine what environmental of the first and second antennae, mandible, sec- factors affected their distribution. In discussing ond maxilla, first through third maxillipeds, first relationships among the Decapoda, Burkenroad pereiopod, telson, uropod, branchiae, and first (1981:261) mentioned that A. scabra does not and second pleopods of a male. Aurivillius (1898) possess "tack-like" sperm. illustrated his A. margaritacea var. claviger as fol- PUBLISHED ILLUSTRATIONS.—The earliest illus- lows: the telson, the third pereiopod, and dorsal tration of this shrimp is that of Marcgrave (1648), and lateral views of the carapace. Gerstaecker who presented a lateral view of the entire animal and Ortmann (1901) figured an animal in lateral (presumably from Brazil), identifying it by the aspect along with views of the antenna, mandible, vernacular name "guaricuru." This drawing was first and third maxillipeds, and first pereiopod. later employed by Piso (1658) and others. Almost Bouvier (1905) presented drawings of the rostrum 114 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY of A. scabra in dorsal aspect and a lateral view of photograph of an entire animal in lateral and the cephalic part of the carapace. The drawing of dorsal views along with drawings of the rostrum, Torralbas (1917) was of a specimen in lateral appendices masculina and interna, and of the view. In his monograph, Bouvier (1925) illus- antennular peduncles. Bonnelly de Calventi et al. trated the eyes, basal segment of the antennule, (1974) provided a photograph of a shrimp in the antennular carina (= ocellar beak), the first lateral view. Leveque (1974) figured the rostrum and second maxillae, dactyls of the fourth and in dorsal aspect as did Alayo (1974). Fryer's fifth pereiopods, and reprinted drawings of the (1977) illustrations include a lateral view of a rostrum in dorsal aspect and a lateral view of the specimen, setal structural features, and photo- cephalic region of the carapace. The latter two graphs of a Dominican animal filter feeding and are mislabeled, for figures 705 and 706 (not 703 of sections through the mouth parts. and 704) correspond to figure 25s of A. scabra in DIAGNOSIS.—Cephalic region comparatively Bouvier (1905). Monod (1967), employing Bou- weakly sculptured except nodose in larger indi- vier's (1925) figures, repeated the latter's error. viduals; much of carapace studded with pile of Illustrations of a specimen of A. scabra, presum- short stiff setae (conspicuous in early post molt ably from the Belgian Congo, by De Man (1925) stages); antennal and pterygostomian spines include the rostrum in dorsal and lateral aspects prominent. Rostrum with margins bearing paired and the telson in dorsal view. preapical angles, latter almost always slightly The most completely illustrated work on Atya produced anteriorly; dorsal surface with rather scabra is that of Villalobos (1943) in which vir- sharp median carina. Ventral margin of second tually the entire exoskeleton is depicted. Oliveira through fifth abdominal pleura with row of scle- (1945) presented drawings of the animal in lateral rotized denticles (occasionally absent on second aspect, a dorsal view of the carapace, the caudal and fifth pleura throughout the range of the area in dorsal and ventral aspects, and the max- species and almost characteristically wanting on illae and rnaxillipeds. Rioja, Ruiz, and Larios the second in members of African populations); (1955) reproduced the lateral view of the entire caudoventral angle of fourth and fifth pleura animal as illustrated by Villalobos (1943). The acute but rarely produced. Sternum of sixth ab- illustrations provided by Parodiz (1960) include dominal segment little more than half as long as the cephalic region in lateral aspect, a lateral wide; compressed median prominence on ster- view of the third pereiopod from the merus dis- num of fifth abdominal segment conspicuous but tally, and the distal part of the firsto r second (not never forming hornlike element overlapping sixth third as labeled) pereiopod. abdominal sternite. Preanal carina acute to Hart (1961b) presented a photograph of a Ja- rounded apically and directed caudoventrally, maican specimen in lateral aspect. A similar pho- apex rarely reaching caudal margin of basal part tograph along with a dorsal view of the rostrum of carina. Telson about 1.5 to 1.7 times as long as and a lateral view of the cephalic region are broad with paired dorsal rows of 5 to 7 spines. recorded by Davant (1963). Chace and Hobbs Antennular peduncle with dorsal surface of prox- (1969) included a lateral view of a male from imal article bearing 1 to 3 sclerotized spinules Dominica showing the color pattern, a dorsal proximal to transverse distal row; penultimate view of the cephalic region, lateral views of the article 1.3 to 1.6 times as long as wide and bearing preanal carina and of the ventral margin of the scattered spinules on dorsal surface. Coxae of second abdominal pleuron, and mesial views of third and fourth pereiopods lacking prominent the distal part of the second pleopod of the male anterolateral spine. Third pereiopod with merus and of the appendix masculina. Rutherford rounded ventrally, about 3 times as long as high, (1971) presented illustrations of the rostrum in ventromesial surfaces of paired articles not touch- dorsal view and of the cephalic region in lateral ing when appendages brought together, and lat- aspect. Bonnelly de Calventi (1974b) included a eral surface bearing irregular rows of prominent NUMBER 364 115

corneous-tipped tubercles; propodus 1.7 to 2.1 nated by Holthuis) as the neotype of Leach's Atys times as long as broad and studded with rows of scaber. sclerotized tubercles on both extensor and flexor NEOTVPIC MALE.—Rostrum (Figure Aba,d) surfaces, tubercles in row on flexor surface mesial with concave lateral margins slightly produced at to median line of article quite or almost contig- base of acumen in angles little less than 90 de- uous; dactyl freely movable and bearing single grees; acumen just overreaching proximal podo- row of usually 6 or 7 tubercles on flexor surface. mere of antennule; dorsal median carina high, REMARKS.—Much of the misunderstanding of rather sharp, excavate dorsally (dipping to or the range and variations in Atya scabra no doubt below level of lateral carina posterior to acumen), resulted from the fact that Leach described the and reaching apex of acumen; ventral carina with species from specimens obtained from an un- single anteriorly directed corneous-tipped tooth. known locality. Despite the existing confusion, Ocellar beak, largely obscured by eyes, tubercu- most previous authors have agreed that the Amer- liform with anterior margin almost vertical and ican atyids with a heavily pubescent body and apex directed dorsally, latter fitting into shallow the third pereiopods studded with large corneous subrostral groove. Antennal spine acute; ptery- tubercles are members of Leach's species. Cloud- gostomian spine slender and acute, no spines ing this general concept of A. scabra, however, is present between them. Carapace (length 29.8 a dry specimen in the British Museum identified mm) devoid of other spines, but surface somewhat as A. scabra and bearing a label on which is nodose dorsally; most conspicuous nodules ar- written "Type." The specimen is clearly identifi- ranged in 1 to 3 subparallel rows extending an- able with Atya gabonensis Giebel and exhibits char- teroventrally, and almost entire surface bearing acteristics that would hardly have been ignored punctations studded with clusters of 2 to 5 short, by the delineators of the illustrations presented erect setae. Conspicuous subcircular obliquely by Leach (1815, 1816) and H. Milne Edwards impressed area present at cephalic end of bran- (1816). The label is therefore strongly suspect. chiocardiac groove, and 2 smaller flat ones situ- As Holthuis (1966:234) has pointed out, New- ated posterodorsal to it, these areas smooth and port (1847:158) noted that there were four speci- lacking punctations and setae. mens of this species "in the cabinets of the British Pleuron of second abdominal segment (Figure Museum, but nothing whatever is known of their 45i) with rounded posteroventral extremity, that habits, or from whence they were obtained." of third, fourth, and fifth weakly produced in Continuing his remarks on the types of this very short subspiniform corneous prominences. shrimp, Holthuis stated: Ventral margins of second through fifth pleura with linear clusters of 14, 17, 11, and 8 corneous In White's (1847:74) list of the Crustacea in the British denticles on right side of body, respectively, and Museum, however, is indicated under Atya scabra "a-d. Mex- ico. From the collection of Dr. Leach." It is impossible 6, 18, 11, and 7 on left; margin of fifth lacking [inasmuch as Leach's specimens lacked data] therefore that prominent setal fringe. Fourth, fifth, and sixth the locality of Leach's type material has been discovered abdominal terga approximately subequal in later and that the material listed by White indeed is the type length and 0.9 as long as telson. Sternum of fifth material. abdominal segment (Figure 45/) with rather Although Holthuis did not designate a neotype large, median, laterally compressed projection for the species, he stated (p. 234): "It seems safe having irregular oblique ventral extremity bear- to restrict the type locality to the Atlantic drain- ing row of 3 very small corneous tubercles; ster- age of Mexico and more accurately to the area of num of sixth about 0.67 as long as broad. Preanal Veracruz." In order that stability in the action of carina with caudally directed apex almost reach- Holthuis be more assured, we are designating a ing level of posterior margin of basal (dorsal) part male specimen in the British Museum from Mis- of carina. Telson injured, but that of male in antla (the location of which is in the area desig- same lot of specimens (Figure 45A) about 1.6 116 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 45.—Alya scabra (all from neotype except h from topotype): a, dorsal view of cephalic region; b, dorsal view of antennular peduncle; c, lateral view of preanal carina; d, lateral view of cephalic region; e, lateral view of second through fifth abdominal pleura; /, preanal carina, sternum of sixth abdominal segment, and median tubercle on that of fifth; g, flexor surface of distal part of fifth pereiopod; A, dorsal view of telson; i, flexor surface of distal part of fourth pereiopod; j, lateral view of third pereiopod; k, mesial view of appendices masculina and interna; /, flexor surface of distal part of third pereiopod. (Scales marked in 1 mm increments.) NUMBER 364 117 times as long as wide, its dorsal surface bearing most reaching distal end of segment, dorsal sur- paired, mesially concave rows of 6 corneous den- face with linear cluster of setae and 3 corneous ticles each and posteromedian tubercle, latter denticles; distal margin studded with row of 7 very slightly overreaching caudal margin of tel- (right) or 6 (left) cornified spinules; penultimate son. segment of peduncle 1.6 times as long as wide, its Proximal podomere of antennule (Figure 45b) dorsal surface bearing 22 small, irregularly ar- with stylocerite bearing acute, corneous tip al- ranged, corneous spinules, dorsodistal margin

FIGURE 46.—Atya scabra (all from male from Rio Sabinas, Estado de Tamaulipas, Mexico): a, dorsal view of cephalic region; b, mesial view of appendices masculina and intcrna; c, lateral view of cephalic region; d, lateral view of second through fifth abdominal pleura; e, lateral view of preanal carina; /, lateral view of third pereiopod; g, dorsal view of telson; A, flexor surface of distal part of third pereiopod. (Scales marked in 1 mm increments.) 118 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY with row of 8; dorsum of ultimate segment of ventromesially, there mostly studded with con- peduncle about 0.5 as long as penultimate podo- spicuous tufts of plumose setae borne in puncta- mere, 3 or 4 spinules on dorsal surface, oblique tions. Propodus entirely tuberculate, those on row of 8 and 2 eccentric distolateral spinules flexor surface arranged in 2 longitudinal series, 5 flanking dorsal base of lateral flagellum; similar of those in mesial row contiguous. Dactyl not rows of 4 (right) and 5 (left) at dorsomesial base fused with propodus, flexor surface with single of mesial flagellum, and 3 or 4 on dorsal surface row of 6 tubercles flanked distally by setal clus- proximal to latter rows. Antenna with lateral ters. spine on basis clearly overreaching stylocerite; Fourth pereiopod with dactyl almost reaching lateral spine on scaphocerite extending to slightly distal end of merus of third; length of merus beyond midlength of ultimate podomere of an- about twice that of carpus, and latter only slightly tennule and falling considerably short of distal shorter than propodus. Fifth pereiopod almost extremity of antennal peduncle; lamella far sur- reaching midlength of merus of third; merus passing latter; flagellum broken but reaching at about 1.5 times as long as carpus, and latter 0.65 least to midlength of second abdominal tergum. as long as propodus. Ornamentation of fourth Third maxilliped extending anteriorly to about pereiopod similar to that of third, except merus same level as antennular peduncle; penultimate with articulated distal ventrolateral spine and segment about 1.5 times as long as ultimate; tip ventral one proximally, and carpus with similar of exopod reaching base of distal fifth of penul- large distolateral spine. Ornamentation of fifth timate podomere of endopod. pereiopod like that of fourth, except merus with First pereiopod reaching only slightly farther 2 articulated ventral spines, and flexor surface of anteriorly than second and attaining level (ne- dactyl with row of many more spinules. glecting terminal setae) of about midlength of Diaresis of lateral ramus of uropod with 19 penultimate segment of antennular peduncle; (right) or 20 (left) articulated corneous denticles bristles composing terminal brush of both pereio- and fixed lateral spine, latter about same size as pods lacking scraping denticles. Third pereiopod spinules. (Figure 45;,/) lacking spines, and, when extended COLOR NOTES.—The most complete notes on anteriorly, overreaching antennular peduncle by the color of this shrimp are those of Chace and distal 3 podomeres; merus with ventromesial mar- Hobbs (1969:63-66) based upon specimens col- gin bowed, 2.3 times as long as high, 1.8 times as lected on the island of Dominica (Figure 47). long as carpus, and 1.6 times as long as propodus; latter almost twice as long as wide and about 0.9 Green Phase: Ground color of cephalothorax and abdo- as long as carpus; distoventral margin of coxa men dark green (chromatophores forming reticulate pat- entire and with conspicuous mesial caudoventral tern), darker dorsally, gradually fading ventrolaterally to olive interspersed with dark cream. Dorsum with broken, prominence bearing corneous tip and studded narrow, median, longitudinal, greenish-cream stripe extend- with tufts of plumose setae. Lateral, dorsal, and ing from anterior part of rostrum to posterior margin of fifth ventral surfaces of merus studded with conspicu- abdominal tergum. (In some larger individuals, dorsal light ous corneous tubercles of which many, if not stripe obliterated.) Three dorsally situated transverse bands most, somewhat flattened distally and bearing of dark forest green as follows: (1) immediately posterior to base of rostrum and extending ventrally almost to level of sharp free edge; plumose setae, either singly or in antennal spine; (2) on posterior margin of carapace, contin- groups, present at distal base of tubercles, some uing onto anterior portion of first abdominal tergum, and setal groups partly encircling tubercles; median extending ventrally on both to level of base of abdominal longitudinal row of setae on lateral surface of pleura; and (3) on anterior half of sixth abdominal tergum podomere well defined; distomesial extremity of extending ventrally to margin. Carapace with dark green podomere produced in broad rounded lobe bear- spot immediately below second band and cream one imme- diately anterior to spot, studded with pile of short golden ing row of 5 (right) or 7 (left) corneous denticles. setae, one to three setae in each punctation. Carpus strongly tuberculate except ventrally and Antennular and antennal peduncles forest green; flagella NUMBER 364 119

FIGURE 47.—Lateral view of Atya scabra (from Chace and Hobbs, 1969). tan to brown. Third maxillipeds and first two pairs of ous and is also from Brazil, has a carapace length pereiopods lavender cream with bright green longitudinal of 29.5 mm; the corresponding length of the lines on ischium and merus; proximal portions of chelae of first two pereiopods lavender, distal portions vermilion; dis- smallest ovigerous female is 7.1 mm. tal setal tufts dark gray proximally fading to beige distally. DISTRIBUTION AND SPECIMENS EXAMINED.—Atya Proximal three podomeres of third, fourth, and fifth pereio- scabra ranges from Liberia to northern Angola, in pods cream with irregular green markings; merus with the islands off the west coast of Africa from the cream-colored base followed distally by broad green band, Cape Verde group southward to Annobon, in the narrower cream one, and distal green one; carpus and West Indies from Cuba and Hispaniola to Cura- propodus of third pereiopods forest green; those of fourth and Fifth cream proximally and dark green distally; dactyls sao and Trinidad, and from Tamaulipas, Mexico, of three orange brown (corneous); tubercles on all three legs to Santa Catarina, Brazil. In Panama it has progressively darker from proximal to distal podomeres. apparently crossed the continental divide, occur- Basal portions of pleopods cream with greenish-tan lateral ring in the Rio Frijoles Basin (Figures 49-52). Its margin; rami tan with brown borders. Uropod light green presence elsewhere on the Pacific slope perhaps proximally, dark green distally with patches of brownish pigment alternating with aqua spots in distal portions of resulted from introductions. both rami; marginal setae golden. Telson light green ante- Records for the known localities are listed be- riorly, dark green posteriorly with brownish spots and brown low. Collections that we have examined are tip. marked with an asterisk if they have been previ- Brown Phase: Pattern essentially identical; colors, how- ously reported and with a dagger if they are ever, range from cream through pale buff to dark brown. reported herein for the first time. Numbers follow- SIZE.—The largest specimen that we have ex- ing the specimens listed are measurements, in amined is a male from Brazil exhibiting a cara- mm, of the carapace length or, if followed by pace length of 46.3 mm, and the largest in African "t.l.," total length. Some listings lack dates and/ waters is also a male with a corresponding length or collectors; these could not be determined. of 34.4 mm. The largest female, which is oviger- LIBERIA: (1) *USNM, St. Paul River, Mt. Cof- 120 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 48.—Atya scabra (from male from Sao Tiago, Cape Verde Islands): a, dorsal view of cephalic region; b, mesial view of appendices masculina and interna; c, lateral view of cephalic region; d, lateral view of second through fifth abdominal pleura; e, lateral view of preanal carina;/, lateral view of third pereiopod; g, dorsal view of telson; h, flexor surface of distal part of third pereiopod. (Scales marked in 1 mm increments.)

fee (Rathbun, 1900:313), Id* (17.3), 22 Mar 1897, GHANA: (1) Environs of Cape Coast between R.P. Currie. (2) fBM> St. Paul River near Handi Elmina and Anomabu (Rutherford, 1971:87,88). / (6°54 N, 10°22'W), 1$ (24.3), 6 Mar 1970, R. CAMEROON: (1) "Etome in Bachen" (Aurivillius, Garms. (3) *RNHL, Sheffelinsville between Mon- 1898:16). (2) Victoria (Balss, 1925:239). (3) Bim- rovia and Marshall (Holthuis, 1966:235), 16* bia River near Dikullu (Monod, 1928:121; (12.2), 1887, J. BGttikofer. (4) fBM, Cavalla 1933:462). (4) stream near Tiko (Monod, River near Nyaake, Grand Gedeh Co (4°51'N, 1933:462). (5) Kienke River near Kribi (Monod, 7°35'W), \6 (23.1), 13 Dec 1970, RG. (5) 1933:462), (6) fRNHL, about 3 km N of Kribi, fUSNM, Liberia Harbel, 64 km inland from 1 juv (3.9), 5 Aug 1964, B. de Wilde-Duyfjes. (7) Monrovia, 19 (17.7), "Smithsonian Firestone *RNHL, pools and waterfalls of the Lobe River Exped. to Liberia." (6) fUSNM, Cavalla River about 9.0 km S of Kribi (Holthuis, 1966:235), 56* at Bolobo, 26 (20.1, 23.3), 1$ (10.2), H.A. Beatty. (6.2-8.9), 6° (5.1-9.8), 3 ovig 9 (7.1-8.8), 23 juv NUMBER 364 121

FIGURE 49.—Distribution of Atya scabra (circled numerals number of localities; see Figures 50-52).

May 1969, R.L.H. Disney. (10) fBM, Blackwater River (4°22'N, 9°47'E), 1

Sao Antonio Valley (Bate, 1888:694). (2) fBM, (Holthuis, 1966:235), 4^(12.1-17.1), 15Feb 1954, Ribeira Picos, 86* (13.9-23.3), 1$ (12.9), C. Stoner; HL. (6) *RNHL, Praia (Holthuis, 1966:235), 20 366* (14.6-25.3), 6$ (11.5-16.6), CS. (3) fUSNM, juv (2.9-4.9), 5-19 Feb 1954, HL. (7) *RNHL, river, 46 (19.7-24.9), CS. (4) fMHNP, Ruisseaux Ribeira Picos, near Picos (Holthuis, 1966:235), descendant de Pics de Antonia au dessus de 500 86 (6.6-24.8), 12$ (6.0-14.3), June 1950, JC. m alt, 116 (19.6-34.4), 1$ (13.9), 10 ovig 9 FERNANDO POO: (Bouvier, 1904:138). *BM, 36 (15.9-20.3), A. Chevalier. (5) *RNHL, Lagoa (10.1-14.4), 3$ (7.0-10.0), 1 ovig 9 (13.5).

FIGURE 51.—Distribution of Atya scabra in the Greater Antilles. NUMBER 364 123

IHLA DO PRINCIPE: (Osorio, 1898:194). (1) Rio Papagaio (Osorio, 1889:139). (2) Rio Banzu (Osorio, 1895b:251). SAO TOME: (Greeflf, 1882:35). (1) Riviere Agua Grande (Osorio, 1887:222). (2) Batipa (Osorio, 1889:137). (3) Obo Vermelho (Osorio, 1889:137). (4) RioQuija (Osorio, 1891a:47). (5) Fleuve Man- uel Jorge (Osorio, 1891b: 140). (6) Rio Gumoela (Osorio, 1892:200). (7) Portinho (Osorio, 1892:200). (8) Rio Agua Ize (Osorio, 1906:150). (9) *RNHL, Bombom (Holthuis, 1966:235), Id (18.1), 19 (13.3). (10) *MHNP, A. Negreiros (Bouvier, 1925:316), 16 (20.3), 1900. (11) fMHNP, Rio de Ouro, 1 ovig 9, 24 Jul 1906. ANNOBON: (Osorio, 1898:194). (1) RioSaoJoao (Osorio, 1895a:249). (2) *RNHL, Crater Lake (Balss, 1914:97), effluent from (01°25'S, 05°36'E), 556* (7.0-25.1), 28$ (6.1-11.2), 16 ovig 9 (7.2-16.7), 20 Mar 1965, Pillsbury sta 278; USNM, 36 (9.0-21.1), 1$ (8.8), 2 ovig 9 (11.0, 19.6). CUBA: (Martens, 1872:135). (1) Rio Almendares at Calabazar, Provincia de la Habana (Bouvier, 1925:317); *USNM, 36 (8.0-19.2), C.F. Baker. (2) Calabazar, Provincia de la Habana (Bouvier, 1909:333). (3) fMHNP, La Habana, Provincia de la Habana, 46 (11.8-17.7), 19 (12.1), 1909. (4) Rio San Vincente near Vifiales, Provincia de Pinar del Rio (Holthuis, 1974:231). JAMAICA: (Ortmann, 1895:410, in error for Newport's Atya occidental^). *USNM, 29 (20.9, 21.0), 1-11 Mar 1884, Albatross. (1) Thomas River 1 mi N of Summerfield, Clarendon Pr (Hart, 1961b:64). (2) *RNHL, Rio Minho at confluence with Pennants River, Clarendon Pr (Hart, 1961b:64), 26 (23.5, 26.8), 9 Apr 1959, C.W. Hart. (3) *USNM, Byndloss Gully, 1.5 mi N of Linstead on road to Ewarton, St. Catherine Pr FIGURE 52.—Distribution of Atya scabra in the Lesser Antilles (Hart, 1961b:66), 46 (18.0-26.2), 19 (18.2), 7 Apr (circled numerals = number of localities). 1959, G. Thomas, CWH. (4) Cane River, St. Andrew Pr (Hunte, 1978:142). (5) Morant River, fBM, Thorlands, \6 (18.0), 19 (18.9), 2 ovig 9 St. Thomas Pr (Hunte, 1978:142). (6) Bugaboo (12.4, 16.3), 17 May 1937, IS. Dominican Republic— River, St. Thomas Pr (Hunte, 1978:142). (7) (1) *USNM, "San Domingo," \6 (16.0), 1878, , Portland Pr, alt 80 m (Hunte, W.M. Gabb. (2) Rio Haina (Bonnelly de Calventi 1978:142). et al., 1973:1338), \6 (30.5), I. Bonnelly de Cal- HISPANIOLA: Haiti— (Kingsley, 1878a:92). (1) venti. (3) *MCZ, Rio Nigua (Bonnelly de Cal- fBM, Hinche, 26 (20.4, 30.1), I. Sanderson. (2) venti et al., 1973:1338) near San Cristobal, Tru- 124 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY jillo Pr, 56* (14.2-18.2), 3$ (10.2-11.5), 5 ovig 9 *USNM, Mannet's Gutter near mouth (Chace (10.1-11.3), PJ. Bermudez. (4) fUSNM, Rio San and Hobbs, 1969:66), 16* (27.8), 1 Apr 1964, Juan, 16* (9.6), S.G. Miller, Jr. (5) Rio Mana HHH; 16* (23.4), 19 Feb 1966, HHH; 26* (22.1, (Bonnelly de Calventi et al., 1973:1338). (6) Can- 23.6), 19 (15.6), 21 Feb 1966, HHH. (3) *USNM, ada Madrigal (Bonnelly de Calventi et al., Mannet's Gutter at upper bridge (Chace and 1973:1338). (7) Rio Isa (Bonnelly de Calventi et Hobbs, 1969:66), 16* (21.1), 13 May 1966, Nichols al., 1973:1338). (8) Rio Jura (Bonnelly de Cal- and Oliver. (4) *USNM, Fond Figues River (to venti, 1974a: 16). (9) Rio Nizao (Bonnelly de Cal- Castle Bruce River) (Chace and Hobbs, 1969:66), venti, 1974a: 16). 16* (11.5) 22 Mar 1964, HHH. (5) *USNM, Pi- PUERTO RICO: *USNM, San Juan Market chelin River below Logge (to Stewart's River) (Gundlach, 1887:131), 16* (18.5), 14 Jan 1899. (2) (Chace and Hobbs, 1969:66), 16* (31.2), 26 Mar *USNM, falls of Rio Grande de Aibonito (Rath- 1964, CWH, HHH. bun, 1901:119), 16* (22.9), 26 Jan 1899, Fish Hawk. MARTINIQUE: (Sharp, 1893:111). fMHNP, La (3) Falls of Rio Grande de Arecibo (Rathbun, Trinite, 36* (27.4-37.5), 29 (17.0, 18.8), 1 ovig 9 1901:119). (4) El Yunque (Rathbun, 1901:119). (17.2), 1916. (5) Trib to Rio Anasco (Rathbun, 1901:119). (6) SAINT LUCIA ISLAND: fBM, Marc Stream, 2 ovig Rio Espfritu Santo at El Verde (Villamil and 9 (11.4, 12.1), 18 Jun 1971, G. Barnish and R.F. Clements, 1976:5). (7) fUSNM, Maricao, creek Sturrock. below town, 3$ (11.0-12.5), 7 Feb 1934, S.F. GRENADA: (1) fRNHL, La Sagesse River, Coats Hildebrand. (8) fUSNM, Rio Maricao, 26* (28.0, Gap, St. Davids, 1 ovig 9 (17.9), 9 Jan 1966, 29.8), 1$ (14.4), 9 Feb 1951, N.T. Mattox. (9) Justin Francis. (2) fRNHL, Beaulieu, 1 ovig 9 fUSNM, (?) Maricao, Rio Grande de Anasco, 16* (16.9), 3 Oct 1965, J.R. Groome. (3) fRNHL, St. (about 20), 21 Feb 1934, SFH. (10) fUSNM, Marks River, Bonaire Estate, 16* (11.5), 25 Jan Juana Diaz, 86* (24.4-33.3), 19 (15.1), 26 Feb 1966, C.A.O. Philips. 1934, SFH. (11) fUSNM, Rio Comerio above S TOBAGO: (Ortmann, 1895:410). *BM, 26* (13.4, dam, 56* (36.5-40.9), 6 Mar 1934, SFH. (12) 14.9), 1 ovig 9 (11.0), P.L. Guppy. (1) fRNHL, USNM, Rio Culebrinas at Rte 13, 100 m S and Frenchman's River at Speyside, 16* (17.3), 19 300 m E of San Sebastian, 46* (18.0-26.4), 2 Jun (10.1), 1 ovig 9 (11.5), 18 Jan 1955, PWH. 1953, H.W. Harry. (13) fUSNM, Rio Cruces at TRINIDAD: *BM, 26* (34.0, 40.0), PLG. (1) Rte 2, 1500 m W of Sabana Grande, 26* (18.4, *MHNP, Port of Spain (Bouvier, 1925:316), 16* 22.7), 16 Jun 1953, HWH. (14) fUSNM, Maya- (31.8), 19 (22.0), 1914, Paul Serre. (2) fMCZ, giiez, 1 ovig 9 (13.3), 20 Jan 1899, Fish Hawk. (15) brook at Maguerepe Bay, 16* (18.0), 7 ovig 9 fUSNM, Rio Guanajibo, about 2,000 m E of San (10.4-13.2), 24 Jul 1937, E. Deichmann. (3) fBM, German, 16* (22.6), 16 Jun 1953, HWH. (16) Mt. Aripo, in small, clear, rocky river, 16* (16.1), fUSNM, stream at Jayuya, 16* (42.2), Spring 19 (20.0), 10 Aug 1937, IS. (4) fBM, Maracas 1954, L.A. Costas. River, 46* (26.1-35.2), 19 (19.5), 2 ovig 9 (18.6, SAINT KITTS: fRNHL, Wingfield, 5 juv 20.3). (3.2-3.5), 30 Jun 1949, P.W. Hummelinck. CURA£AO: fRNHL, Santa Cruz Plantation near MONSERRAT: fBM, river at Monserrat, 46* Santa Cruz, 19 (9.7), 1 ovig 9 (11.0), 11 Jan 1957, (20.4-26.1), 1$ (21.0), 2 ovig $ (16.4, 19.8), F.H. L.B. Holthuis. Mansell. MEXICO: (Guerin-Meneville, 1829-1844:15). GUADELOUPE: Riviere Bras David (Leveque, *RNHL, 2 dry spec (29.5, 34.8). *USNM, 16* 1974:42). (35.1). Tamaulipas— (1) fMCZ, Rio Frio, 16 km DOMINICA: (Oliviera, 1945:177). *USNM, 26* N of Mante, 26* (15.0, 18.3), 21 Dec 1939, W.A. (17.5, 22.3), 19 (18.0). (1) *USNM, trib to Layou McLane. (2) *USNM, Rio Sabinas (to Rio Ta- River across from Clarke Hall (Chace and Hobbs, mesi) at Storm's Ranch, km 619 Carretera Inter- 1969:66), 16* (20.1), 4 Feb 1964, HHH. (2) americana, 4.8 km NE of Gomez Farias (Darnell, NUMBER 364 125

1956:131), 36* (22.7-25.5), 1 ovig 9 (16.2), 27 Apr 18 Apr 1963, HL. (7) fUSNM, Rio Lagarto 1950, M. Darnell. Veracruz—(I) fUSNM, Santa between La Campana and Bejuco, 26* (9.6, 12.2), Maria, 26* (18.0, 19.5), 14 Feb, Nelson and Gold- 29 (8.4), 1 ovig 9 (9.5). (8) fLGA, "Atlantic man. (2) fUSNM, Orizaba, 1 ovig 9 (14.4). (3) Pipeline Road," 26* (28.6, 31.4), 1973, L.G. Abele *BM, Misantla (Wiegmann, 1836:145), 36* and Robinson. (9) fLGA, stream 20-23 km N of (21.9-29.8), 1 ovig? (18.6), 1888, F.D. Godman. El Llano on Carti Hwy, 16* (11.3), 30 Mar 1973, (4) Rio Necaxa near Coyutla (Villalobos, R.L. Dressier. (10) fLGA, Rio Frijoles, 19 (19.1), 1943:11). (5) Rio Papaloapan (Villalobos, 14 May 1975, Kramer; 16* (25.1) Aug 1974. 1959:328). (6) Tapalapan, Santiago Tuxtla COLOMBIA: (1) Rio Carabali, affluent of Rio (Hobbs, 1971:27). (7) fBM, Hacienda El Hobo, Chagri, Darien (Bouvier, 1904:138). (2) Santa Rio de los Bobos, 16* (16.6), 1 ovig 9 (21.1), Nov Marta (Pearse, 1915:551). 1897, P. Geddes. Oaxaca— (1) *MHNP, Rio Cha- VENEZUELA: (1) Rio Macuto near La Guaira calapa near Pochutla (Bouvier, 1925:316), 16* (Nobili, 1897:5). (2) fUSNM, freshwater rocky (22.3), 1903, Diguet. (2) *MHNP, Valle Nacional branch at Macuto (cited by Rathbun, 1901:119, (Bouvier, 1904:138), 36* (23.6-25.5), 4$ (20.1- as "Venezuela"), 56* (13.5-25.1), 1 Aug 1900, 21.4). Lyon and Robinson. (3) San Esteban (Bouvier, GUATEMALA: (1) *USNM, Rio Gualan at 1904:138). (4) *MHNP, Naricual (Bouvier, Gualan (Bouvier, 1925:317), 36* (17.5-33.5), 1$ 1904:138), 56* (20.0-24.3), 19 (15.6), 1 ovig 9 (20.0), 1 ovig 9 (injured), 4 Jun 1909, C.C. Dean. (20.8), 1885. (5) fRNHL, Rio Tuy, Santa Teresa (2) fRNHL, in drinking water on KNSM ship del Tuy, Estado de Miranda, 19 (18.6), Feb 1949, Breda, water last filled at Puerto Barrios, 1 juv G. Marcuzzi. (6) fUSNM, Rio Cumboto near (3.0), March 1966, J. Geyskes. mouth, 2 km NW of Ocumare, 16* (7.3), 5 May HONDURAS: (1) fUSNM, Lancetilla, 16* (21.1), 1939, F.F. Bond. (7) Rio Manzanares, Cumana, 14 May 1946, A.F. Carr, Jr. (2) fUSNM, Rio Estado de Sucre (Davant, 1963:100). Guayabal, trib to Rio Comayagua, Dept de BRAZIL: (1) Rio Prangi, Estado do Pernambuco Cortes, alt about 100 m, 26 (28.9, 33.3), 22 May (Oliveira, 1945:177). (2) Rio Serinhaem, Estado 1948, A.C. Chable. (3) fUSNM, Lago de Yojoa, do Pernambuco (Lemos de Castro, 1962:50). (3) 16* (32.2), 7 Nov 1954, S.Y. Lin, Jr. fUSNM, Blumenau, Estado do Santa Catarina, NICARAGUA: fUSNM, Waspuc River at Musa- 16* (46.3), 1 ovig 9 (29.5), Feb 1926, Paula Paul. was, 36* (20.0-32.5), 2 ovig 9 (13.5, 16.7), Oct ERRONEOUS OR PROBABLY ERRONEOUS LOCALI- 1955, B. Malkin. TIES.—There is reason to question the following COSTA RICA: fBM, Salamanca, 16* (20.1), March localities as noted for each. 1895, HP. AUSTRALIA: Victoria (Bouvier, 1904:138). Two PANAMA: *LGA, 96* (14.1-23.9). (1) Atlantic specimens reported by Bouvier, if correctly iden- side (Doflein, 1900:127). (2) *USNM, Ri'o tified, almost certainly bore erroneous data, for Chagres Basin, Isla Barro Colorado (Allee and no member of the genus, as defined herein, has Torvik, 1927:67), 86* (6.2-15.0), 20 ovig 9 since been reported from the continent. (9.2-11.4), Aug 1962, H. Loftin. (3) fUSNM, NEW CALEDONIA: (Ortmann, 1890:465). The large river on mainland opposite Isla Multatupo, three syntypes of Atya margaritacea A. Milne-Ed- 66* (6.0-14.1), 49 (5.0-9.8), 1 ovig 9 (9.4), 2 Dec wards (1864), considered by Ortmann to belong 1962. (4) fUSNM, San Bias area, creek just W of to A. scabra, constitute the basis for this erroneous Puerto Obaldia, 76* (6.5-8.2), 4 ovig9 (10.1-10.4), record. As pointed out elsewhere, that the speci- 9 Feb 1963, HL. (5) fUSNM, San Bias area, mens were collected on New Caledonia is highly small river opposite Isla Ustupo (Portogandi), 16* improbable. (10.8), 19 (11.6), 1 ovig 9 (9.3). (6) fUSNM, ROLAS ISLAND: (Rathbun, 1900:314). Appar- Chiriquicito, backwaters and banks of Rio Guar- ently Rathbun erred in recording Rolas Island as umo, 26* (16.8,18.3), 19 (15.4), 3 ovig 9 (9.2-13.7), having been cited to be frequented by A. scabra 126 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY by GreefT, and her error was repeated by Balss punctata from the insular and continental Ameri- (1914:98). can Atya scabra (including A. mexicana). Further- BAJA CALIFORNIA: (Holthuis, 1951:25). Consid- more, the nature of the variations that have been ering Atya rivalis (= A. margaritacea) to be a syn- reported to serve as a means of separating them, onym of A. scabra, Holthuis cited Lower California even were they reliable, seems to us to be so minor as marking the northern limit of the range of the that they are hardly noteworthy. Most of the latter on the west coast of the Americas. specimens that we have examined from Africa WESTERN MEXICO: (Stimpson, 1857:498). We and the eastern Atlantic islands lack corneous have not located the specimens in the Smithson- denticles on the ventral margin of the second ian collections on which Stimpson based this abdominal pleuron, but individuals in several lots record. Perhaps it is a valid one, for the lot cited possess one or more of them. The reverse is true here from Rio Chacalapa, Oaxaca, contains a of specimens from the Americas (including the male of this species rather than one of A. margar- West Indies). Comparisons of proportions of sev- itacea as we had anticipated. eral features of the antennules and pereiopods NICARAGUA: (Rathbun, 1900:313). The Nicara- resulted in the discovery of no feature that could guan record was based upon the assumed synon- be associated with a restricted part of the range. omy of A. rivalis with A. scabra. The types of the The setal pile so conspicuous in recently molted former were from Nicaragua. A new locality on individuals is sometimes exceedingly sparse. The the eastern versant (see above) has established the rostral margins may be subparallel or rather presence of A. scabra in the country. strongly concave; the dorsal carina varies from COSTA RICA: (Rathbun, 1900:313). The locality being moderately high and rounded to high and record for Costa Rica by Rathbun, and elabo- sharp and frequently concave, sloping over the rated upon by Bouvier (1925:316), was based on main body of the rostrum to the level of the a collection of A. margaritacea from Rio Platanales, lateral carinae; toward the apex, it may be Mt. Chiriqui, which is in Panama. To our knowl- smooth or scalloped but terminates at, or almost edge, the first valid record of the occurrence of at, the tip of the acumen. The width and depth the species in this country is the new one cited of the furrows flanking the median carina are also herein. quite variable but, in general, are rather broad VENEZUELA: (Rathbun, 1900:313). Rathbun's and shallow in small individuals, becoming recording the presence of A. scabra in the Orinoco deeper and the fundal area narrower (as a result was based on the assumed synonomy of Koelbel's of thickening of the basal parts of the lateral and A. sculptilis {— A. gabonensis) from the Orinoco median carinae) in larger, presumably older, in- River with A. scabra. Apparently she was unaware dividuals. The ventral carina may bear one or no of the locality "Rio de Macuto" that had been spine along the anterior half. reported by Nobili (1897:5). The width of the penultimate podomere of the ARGENTINA: fBM, Buenos Aires, 1<3 (22.5), Mar antennule ranges from only slightly less than the 1892, H.P. Hirt. Within the lot containing this length to a little less than half as long. The basal specimen of A. scabra were two specimens of A. podomere of that appendage bears none to three margaritacea. Inasmuch as the latter has been corneous spinules dorsally, and the premarginal found elsewhere only in the Pacific watershed spinules on the dorsal surface of the penultimate and has not been collected in other localities with and ultimate podomeres vary both in number A. scabra, the source of the specimens contained and position, the more laterally situated ones on in this lot must be questioned. the penultimate segment sometimes occurring in VARIATIONS.—Try as we may, we have found a linear series. no character or group of characters that enable The merus of the third pereiopod ranges from us to distinguish the African Atya sulcatipes, A. about 2.0 to 2.7 times as long as high; the pattern margaritacea var. claviger, and the American A. of distribution of the tubercles on its lateral sur- NUMBER 364 127 face follows somewhat that illustrated in Figures 1844:15), H. Milne Edwards (1837:348), and oth- 45/, 46/, 48/, but the details are apparently never ers apparently gave some subsequent authors the identical. In small individuals, a single ventrolat- impression that A. scabra occurred in salt water, eral spine occurs near the distal end of the podo- for Stimpson (1857:498) stated that "Milne-Ed- mere but is absent or reduced to a vestige in wards considers A. scabra to be a marine form, but virtually all individuals with a carapace length there is great doubt that any species of the genus greater than 20 mm. Whereas the number of is found in the sea." Gundlach (1887:131) found tubercles on the flexor surface of the propodus of this shrimp in a tributary to Rio Anasco in the the third pereiopod may vary, some of those interior of Puerto Rico. Osorio (1887-1906) re- forming the row on the mesial side of the median ported its occurrence in several streams in the line are contiguous or overlap (Figures 45/, 46h, Cape Verde Islands and on other islands off the 48A, 53). The spines on the merus and carpus of West African coast at altitudes of 200 (1892:200) the ambulatory appendages in many, if not most, to 300 m (1891b: 140). Rathbun (1897:44) reiter- individuals with carapace lengths under 20 mm ated its presence in "Fresh waters of tropical are dispersed as follows: the merus of the third America and Cape Verde Islands," and Aurivil- pereiopod exhibits one distal ventrolateral spine, lius (1898:16) reported his Atya margaritacea var. and the carpus lacks spines; the merus of the claviger from "brooks." Pearse (1915:551) noted fourth pereiopod possesses in addition to one that in the area of Colombia in which he was spine corresponding to that on the third another exploring, "This peculiar shrimp was quite com- proximoventral to it, and the carpus bears one mon in the swift streams flowing among bowlders distal ventrolateral spine; the spination of the [sic] over sandy bottoms near 'La Rosa.' It was fifth pereiopod differs from that of the fourth in found among or under bunches of dead leaves the presence of an additional spine on the ventral which had accumulated against sticks or stones." surface of the merus. Whereas spines are lacking Allee and Torvik (1927:67) stated that in Shan- on the third pereiopod in virtually all individuals non Creek on Isla Barro Colorado, A. scabra was with a carapace length of as much as 20 mm (and limited to the lower 24 pools. Villalobos in some others that are smaller), in the largest (1943:11), in describing the locality on the Ne- specimen available the distolateral one is present caxa River from which his specimens were ob- on both the merus and carpus of the fourth and tained, stated that it was situated 100 km from fifth; in some of the larger individuals, at least the sea, at an elevation of about 800 m in a one of the ventral spines is present on the merus tropical area surrounded by . The of the fifth, and in some individuals two are stream at this site is shallow, and the shrimp live present on the merus of both the fourth and fifth under the rocks in places where there are riffles. legs. A collection made in this locality with a U-shaped ECOLOGICAL NOTES.—Ten years after this net placed below the rocks as they were turned shrimp had been described from an unknown yielded 127 specimens in less than a half hour. locality, Guilding (1825:338) reported its occur- Darnell (1956) conducted a populational study of rence "in incredible numbers in the mountain A. scabra in Tamaulipas, Mexico. In describing streams" of Saint Vincent. As expressed in the the locality (pp. 131-132), he stated: "Review of Literature" above, Guilding probably did not distinguish this shrimp from the more The habitat of the shrimp was a turbulent riffle, 0-1 meter commonly occurring Atya innocous. Newport in depth, with a rock and boulder bottom.. . . Filamentous (1847:159) noted that Atya sulcatipes had been algae {Lyngbya and Oscillatoria) made up the only endogenous found "300 feet above the level of the sea on San vegetation. This, however, constituted less than 5 percent of the total vegetation in the riffle, the remainder being com- Nicolas, Cape Verde Islands." The reported oc- posed of exogenous river borne branches, twigs, leaves, and currence of members of the species from "les cotes seedpods. Vegetation from the two sources together du Mexique" by Guerin-Meneville (1829- amounted to 45.1 grams per square meter of riffle bottom. 128 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

FIGURE 53.—Atya scabra, variations in flexor surface of distal part of third pereiopod (numbers in parentheses — carapace length in mm): a, Cavalla River, Liberia (6\ 23.1); b, Rio Haina, Dominican Republic (

In addition to the shrimp, an abundant and varied gastropod been found in only a few localities, those at and insect fauna was present.... elevations between 17 and 117 meters where it Throughout six discontinuous months of field work dur- ing the winter, spring, and early summer no specimens of "frequents cascading reaches and riffles of small this shrimp were observed outside the riffle during daylight streams ... [secreting] itself among the stones hours. At night, however, they were occasionally noted on over and among which water is rushing. Although rocky bottoms in adjacent shallow backwaters, particularly it must travel through pools in moving from one in the vicinity of weed beds. congenial niche to another, it was neither ob- Juveniles and some adult females were found served nor collected in a pool during this study" along the edges of the riffle. An examination of (p. 66). stomach content revealed that 91.3% (by volume) Hart and Hart (1969) recorded chemical and consisted of unidentified detritus, 6.9% of plant physical data collected in two seasons (March remains, and 1.0% remains. Among and November) from a station in Dominica from the plants were diatoms, two of which he attrib- which Atya scabra was taken; Hunte (1978) pro- uted to upstream sources, "Spirogyra sp., desmids, vided additional data from two stations in Ja- and epiphytic diatoms" were thought to have maica. The data from the two localities are sum- originated in weed beds, and Oscillatoria sp. in marized in Table 1. "the local riffle" (p. 135). The small amounts of Villamil and Clements (1976:5) believed A. Oscillatoria and Lyngbya present in the stomachs scabra "to be the most ecologically restricted and suggested to Darnell that the shrimp "were not most specialized member of the crustacean fauna feeding from the uppermost rocks but remained of the area [upper Rio Espiritu Santo at £1 Verde, at the lowermost levels, consuming bits of detritus Puerto Rico]. The females were collected only and other incidental material which filtered from rocky rifflesan d the males only from crevices through to the bottom of the riffle" (pp. 135, between rocks over which a fast current was 136). Few insect parts were present, but among flowing." the arthropod remains were parts of Atya scabra. Fryer (1977) also found A. scabra to be compar- In collecting Potimirim mexicana (De Saussure, atively rare on Dominica and limited to habitats 1857) in the littoral zone of Rio Papaloapan, similar to those mentioned by Chace and Hobbs. Veracruz, Villalobos (1959:328) found juvenile He observed that an adult specimen kept in an specimens of Atya scabra and members of the genus aquarium "by day at least ... had more retiring Macrobrachium to be abundant among the sub- habits than A. irmocous, making few appearances merged roots of the willow (Salix humboldtiand), from the stones under which it took refuge, save which grows along the banks of the stream. He to filter ..." (p. 72). He found members of the also reported discovering that Atya scabra spends species to feed primarily by filtering the water the early stages of its development near the coast, with the aid of its first and second pereiopods. later emigrating up river, becoming dispersed in The entrapped "food is transferred and manipu- areas far removed from the river mouths. He lated by the extremely complex oral machinery stated further that such juvenile forms have gen- ... one of the most elaborate parts of this is a eral similarity to individuals of Potimirim mexicana, teaselling device in which components of the and that in order to distinguish between them a maxillae and first maxillipeds participate" (p. certain knowledge of the distinctive features of 58). The feeding mechanism is described by Fryer the two genera is needed. (pp. 98-104). In their ecological classification of the fresh- Hunte (1978) reported the results of his survey water decapods of the West Indies, Chace and of the freshwater shrimps on the island of Ja- Hobbs (1969:33) categorized A. scabra as a "typi- maica; in each stream he made collections at high cally freshwater species that invades marine hab- and low altitudes. Atya scabra appears to have itats or have, or probably have, marine larvae." been found at altitudes in the four localities cited They reported that on Dominica this shrimp had of 335, 152, 380, and 380 meters (the latter two 130 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

TABLE 1.—Chemical and physical data for one station on the Pichelin River, Dominica (from Hart and Hart, 1969), and two stations in Jamaica, the Cane and Morant rivers (from Hunte, 1978), from which Alya scabra was taken

Pichelin River Cane River Morant River

Mar 1964 Nov 1966

pH 7.7 7.6 7.5 7.7 Temperature (°C) 24.0 24.0 22.8 22.3 Dissolved oxygen (ppm) 8.0 7.9 8.5 8.7 Alkalinity (ppm) 42.0 58.6 - CO2 (ppm) 1.6 3.7 - Cl (ppm) 13.6 13.6 - Total hardness (ppm) 36.0 36.6 - Calcium hardness (ppm) 22.2 25.0 - Magnesium hardness (ppm) 13.8 11.6 - SO4 (ppm) 6.6 6.4 - Ca (ppm) 8.9 10.0 - Mg (ppm) 3.4 2.8 - SiC>2 (ppm) - 68.2 - Specific conductivity - 127.3 - (micromhos/cm)

NO3 (ppm) - - 0.008 0.265 PO4 (ppm) - - 0.384 0.362 the maximum altitude at which collections were substrate exists, although some adult females and made); however, Hunte (p. 139) stated that this juveniles may frequent pools adjacent to riffle shrimp was found at lower altitudes. Classifying areas; young shrimp may also be found among it as a high-gradient species, he also subjected submerged roots of shoreline plants. The altitude specimens to temperatures of 10°, 15°, 35°, and at which this shrimp has been collected ranges 40° C and recorded the survival time. They were from sea level to 800 meters. as follows: 7.7 (±.07), 31.3 (±2.7), 26 (±7.6), and Should the larval requirements prove to be 4.7 (±2.1) hours, respectively. The temperature similar to those of A. lanipes and A. innocous re- of the water in two of the localities where the ported by Hunte (1975, 1979b), then the first shrimp was collected was 22.8° and 22.3° C. molt after hatching must occur in estuarine con- Three specimens were placed in an unaerated ditions or at least in brackish water. aquarium maintained at 25° C. The time of LIFE HISTORY DATA AND POPULATION STRUC- death ranged from 4.2 to 8.0 hours (mean 5.6 TURE.—The earliest report of ovigerous females hours) when the oxygen tensions reached a level accompanied by date of observation is that of of a little less than 1.5 to 0.5 mg/1. The oxygen, Pearse (1915:551), who observed such females NO3, and PO4 concentrations and pH at two of "often during the first week in August" in Santa the localities where the shrimp was found are Marta, Colombia. Villalobos (1943) sampled a noted in Table 1, herein. Salinities of the sea near population of A. scabra inhabiting Rio Necaxa at the mouths of the two streams were 31.9%o and Coyutla, Veracruz, Mexico, during March. His 28.3%o. catch included 100 males and 27 females that There is general agreement among almost all exhibited a mean carapace length of 28.9 (total of the observers that Atya scabra frequents com- length, 68.5) mm and 19.0 (50.0) mm, respec- paratively shallow (records include no more than tively. In November, he obtained about 400 ad- one meter), swiftly flowing water where a rocky ditional individuals among which there were NUMBER 364 131 about 10% more males than were present in the Chace and Hobbs (1969:66) reported collec- collection made in March. He measured 100 tions on Dominica of 15 males and four females males that had a mean carapace length of 29.2 of which one among the latter, collected on 29 (total length, 70.0) mm; some were "individous January, was ovigerous. Leveque (1974) observed gigantes" having a carapace length as great as ovigerous females on Guadeloupe in June. Bon- 38.8 (94.5) mm; smaller adult individuals had nelly de Calventi (1974b:40) reported the occur- corresponding lengths of 23.2 (59.9) mm. Villa- rence of ovigerous females in March, April, May, lobos noted that his collections contained more June, July, September, October, and November. males than females. Ovigerous individuals were The only information on longevity is that of found in both March and November. Fryer (1977:72), who reported maintaining an In studying a population of Atya scabra in the adult individual in an aquarium with several Rio Sabinas in southern Tamaulipas, Mexico, specimens of A. innocous for 2.5 years. Darnell (1956) found the ratio of males to females Summarizing the data presented here, oviger- to be 61:39. He also reported the males to be ous females have been found in African waters in larger: the average total length was 68.5 mm with March, May, July, and August and in American a maximum of 94.5 mm, and the average length waters in every month of the year. The smallest of the females 50.0 mm. Of 23 females collected and largest ovigerous females that we have ex- in late April, four were ovigerous; their total amined have carapace lengths of 7.1 and 29.5 length and number of eggs (in parentheses) car- mm, respectively. ried were as follows: 57.5 mm (8000 estimated), COMMON NAMES.—The following common 50.5 mm (1111), 47.5 mm (1865), and 43.0 mm names have been applied to Atya scabra. (746). The smallest female with ovarian eggs Bomingomo (in Batanga), Cameroon (Monod, 1928:206) "large enough to be considered mature" had a Bouc on Martinique (Pinchon, 1967:161) total length of 39.0 mm. Two females of compa- Cacador in Guadeloupe (Holthuis, 1980:70) and in Domin- rable size had "what appeared to be recently ica; also applied to A. innocous in Guadeloupe (Leveque, spent ovaries" (p. 137). Basing his opinion on the 1974:42) and in Dominica varied stages of development of the ovarian eggs, Camacuto in Venezuela (Davant, 1963:42) he expressed the belief, formerly stated by Villa- Camarao on Sao Tome (Osorio, 1889:139) Chacales (probably derived from the Nauatl word "cha- lobos, that "breeding takes place during most of calli," according to Villalobos) in Veracruz, Mexico (Vil- the year" (p. 137). The average length of the eggs lalobos, 1943:12) was 0.84 mm. Ovarian eggs of 13 nonovigerous Conca, Cruca, or Camarao de pedra in northern Brazil females collected at that time had average lengths (Fausto Filho, 1968:28) ranging from 0.38 to 0.87 mm. He suggested that Coruca in Pernambuco, Brazil (Holthuis, 1980:70) a logarithmic relationship exists between body Crevette gros-doigt in Cameroon (Monod, 1928:458) Dikuta (in Bassa Bania), Cameroon (Monod, 1928:206) size and number of undeveloped eggs carried on Ekusa (in Soubou), Cameroon (Monod, 1928:206) the abdomen of the female. Assuming this to be Guabaras on Cuba (Gundlach, 1887:131) factual, he calculated that the female with a total Guaricuru in Brazil (Marcgrave, 1648:187) length of 50.5 mm should have carried about Inze on Annobon (Osorio, 1895a:249) 3000 eggs instead of 1111 and concluded that Janga or Jonga on Jamaica; apparently applied to any freshwater shrimp on the island (Hart, 1961a:3) "about 60 percent mortality is indicated for a Langostino in Dominican Republic (also applied to A. inno- group of eggs which had reached the eye pigment cous; Bonnelly de Calventi, 1974b:40) stage" (p. 137). Langosta in Dominican Republic (also applied to A. innocous The dimensions of the eggs carried by speci- and A. lanipes; Bonnelly de Calventi, 1974b:40) Mobengomo (in Douala), Cameroon (Monod, 1928:205) mens from the Cape Verde Islands and Venezuela Patua in Dominican Republic (Bonnelly de Calventi, were reported by Bouvier (1925:316) to be 1974b:40) 0.58-0.64 X 0.40 and 0.65 X 0.37 mm, respec- Pepeluna in Dominican Republic (Bonnelly de Calventi, tively. 1974b:40) 132 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Sutu-feelee or "Bitter crayfish" in Africa (Rathbun, stated that Atya scabra were caught for the market 1900:313) on Saint Vincent, and Gundlach (1887:131) re- Vieja in Dominican Republic (Bonnelly de Calventi, ported their being in the market in San Juan, 1974b:40) Viejitas in Cuba (also applied to Atya innocous) (Alayo, Puerto Rico. They were mentioned as being "ex- 1974:22) cellent for eating" by Johnston (1906:843) in Zapata in Dominican Republic (Bonnelly de Calventi, treating the fauna of Liberia. In their list of the 1974b:40) shrimps and prawns of economic importance, According to Holthuis (1980), the FAO names Holthuis and Rosa (1965:9) reported that they for the American Atya scabra are: Camacuto were fished in the following: Mexico, Costa Rica, shrimp (English), Saltarelle camacuto (French), Panama, Nicaragua, British Caribbean Federa- Cameron camacuto (Spanish). For the African tion, Cuba, Haiti, Puerto Rico, Netherlands West representatives (formerly assigned to Atya sulca- Indies, and Venezuela. (Their Peruvian citation tipes), the FAO names are: Ekusa shrimp (Eng- is for Atya margaritacea.) Chace and Hobbs lish), Saltarelle ekusa (French), and Cameron (1969:45) also noted that they were eaten on ecusa (Spanish). Dominica. One of us (CWH) was informed that ECONOMIC IMPORTANCE.—Guilding (1825:338) this shrimp is a food item on Jamaica. Literature Cited

(Citations marked by an asterisk (*) were not seen by the authors, but references to Atya in them were supplied by Lipke B. Holthuis)

Abele, Lawrence G. "Valdivia" 1898-/899, 20:217-315, figures 1-75, 1975. A New Species of Freshwater Shrimp (Genus Atya) plates 20-28. from the Pacific Drainage of Panama. Proceedings 1955. Decapoda, VI: Okologie. In H.G. Bronn's Die Klas- of the Biological Society of Washington, 88(6) :51-58, sen und Ordnungen des Tierreichs, 5 (part 1, book 7, figures 1, 2. number 10): 1285-1367, figures 1043-1069. Abele, Lawrence G., and Nina Blum Bate, C.S. 1977. Ecological Aspects of the Freshwater Decapod 1888. Report on the Crustacea Macrura Collected by Crustaceans of the Perlas Archipelago, Panama. H.M.S. Challenger during the Years 1873-76. In Biotropica, 9(4): 239-252, figures 1-4. Report on the Scientific Results of the Voyage of H.M.S. Alayo D., Pastor Challenger during the Years 1873-76, 24: xc + 942 1974. Gui'a Elemental de las Agua Dulces de Cuba. pages, 76 figures, 157 plates. Torreia, new series, 37: 79 pages, 19 plates. Beatty, Harry A. Albert, Friedrich 1968. A Checklist of Crustacea of St. Croix, Virgin 1883a. Das Kaugeriist der Dekapoden. Zeitschrift fir Wis- Islands. Journal of Agriculture of the University of Puerto senschaflliche Zoologie, 39:444-536, figures 1, 2, Rico, 52(3).-261-268. plates 29-31. Bonnelly de Calventi, Idelisa * 1883b. Ueber das Kaugeriist der Makruren. 80 pages, 2 figures. 1974a. Los Crustaceos de la coleccion del Centra de In- Gottingen: Dieterich'schen Univerersitat-Bucher- vest igaciones de Biologia Marina de la US AD. In uckerei. Estudios de Biologia Pesquera Dominicana. Pub- Allee, W.C., and Magnhild Torvik licaciones de la Universidad Autonoma de Santo Domingo, 1927. Factors Affecting Animal Distribution in a Small 159:13-34. Stream of the Panama Rain-Forest in the Dry 1974b. Camarones de no de Republica Dominicana: Season. Journal of Ecology, 15:66-71, figures 1, 2. Manual para su idemificacion. In Estudios de Audouin, C, and H. Milne Edwards Biologia Pesquera Dominicana. Publicaciones de la 1829. Inconographie des Annelide's, Crustace's, Arachnides et Universidad Autonoma de Santo Domingo, 159:35-64, Myriapodes, ou collection de figures repre'sentant ceux de figures 1-17. ces animaux qui peuvent servir de types pour chaque Bonnelly de Calventi, Idelisa, Sophie Jakowska, and Cecilia famille, avec des details anatomiques, gravees sur cuivre; de la Cruz Matos accompagnee d'une explication des planches, et faisant le 1973. Fresh Water Shrimps ( and Atyidae) complement du Reseume d'Entomologie. 24 pages, 48 of the Dominican Republic and Their Nutritional plates. Paris. Value. American Zoologist, 13(4): 1338. Aurivillius, Carl W.S. Bonnelly de Calventi, Idelisa, Nirva Nunez de Ricart, Sophie 1898. Krustaceen aus dem Kamerun-Gebiete. Bihang Jakowska, and Josefina Perez de Inchaustegui Till Kungliga Svenska Vetenskaps-Akademiens Handlin- 1974. Composicion quimica de los mariscos de mayor gar, 24(IV)(1):3-31, plates I-IV. consumo en Republica Dominicana. In Estudios Balss, Heinrich de biologia pesquera Dominicana. Publicaciones de 1914. Decapode Crustaceen von den Guinea-Inseln, la Universidad Autonoma de Santo Domingo, 159:103- Siid-Kamerun und dem Congogebiet. In Ergebnisse 131, figures 1-27. der Zweilen Deutsche Zentral-Afrika Expedition 1910/ Boone, Lee 11 unter Fuhrung Adolf Friedrichs, Herzog zu Mecklen- 1931. A Collection of Anomuran and Macruran Crus- burg, 1 (Zoologie) :97-108, figures 1-12. tacea from the Bay of Panama and the Fresh 1925. Macrura der Deutschen Tiefsee-Expedition, 2: Waters of the Canal Zone. Bulletin of the American Natantia, Teil A. In Wissenschaftliche Ergebnisse der Museum of Natural History, 53:137-189, figures 1- Deutschen Tiefsee-Expedition auf dem Dampfer 23.

133 134 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Bouvier, E.L. meira Classe de Sciencias Matematicas, Physicas e Na- *1889. Le Systeme nerveux des Crustaces Decapodes et turaes, new series, 4(part 1, number 7): 17 pages, 2 ses rapports avec l'appareil circulatoire. Annales de plates. Sciences Naturelles (Paris), Zoologie, series 7, 6:73- Carvacho, A., and C. Carvacho 106, plate 7. 1976. Une Cle illustree pour la determination des Crev- 1904. Crevettes de la famille des Atyides: Especes qui ettes d'eau douce de la Guadeloupe. In Nouvelles font partie des collections du Museum d'Histoire agronomiques des Antilles et de la Guyane, 2(3):211- Naturelle. Bulletin du Museum d'Histoire Naturelle 219, 2 plates. [Mimeographed.] (Paris), 10:129-138. Cendrero, Luis, editor 1905. Observations nouvelles sur les crevettes de la fam- 1971. Zoologia Hispanoamericana: Invertebrados. xxi + 1151 ille des Atyides. Bulletin Scientifique de la France et de pages. Mexico City: Editorial Pornia, S.A. la Belgique, 39:57-134, figures 1-26. Chace, Fenner A., Jr. 1906. Sur une petite collection de Crustaces (Decapodes 1972. The Shrimps of the Smithsonian-Bredin Carib- et Stomatopodes) recueillis par M. Charles Gra- bean Expeditions with a Summary of the West vier a lile San Thome (Afrique occidentale). Bul- Indian Shallow-Water Species (Crustacea: Deca- letin du Museum d'Histoire Naturelle (Paris), 12:491- poda: Natantia). Smithsonian Contributions to Zoology, 499. 98: x + 179 pages, 61 figures. 1909. Les crevettes d'eau douce de la famile des Atyides In prep. Report of the Carideans of the Albatross Phil- qui se trouvent dans ltle de Cuba. Bulletin du ippine Expedition, 1907-1910. Museum d'Histoire Naturelle (Paris), 15:329-336. Chace, Fenner A., Jr., and Horton H. Hobbs, Jr. 1925. Recherches sur la morphologie, les variations, la 1969. The Freshwater and Terrestrial Decapod Crusta- distribution geographique des crevettes de la fam- ceans of the West Indies, with Special Reference ille des Atyides. Encyclopedic Entomologique, series A, to Dominica. United States National Museum Bulletin, 4: 370 pages, 716 figures. 292: 258 pages, 76 figures. Bouvier, E.L., and P. Lesne Chapa, H. 1901. Sur les Arthropodes du Mozambique et de San * 1964. Lista preliminar de los Crustaceos existentes en el Thome offerts au museum par M. Almada Ne- Laboratorio Central del I.N.I.S.P. Trabajos de Di- greiros. Bulletin du Museum d'Histoire Naturelle vulgacion, Instituto Nacional de lnvestigaciones Bioldgico- (Paris), 7:12-15 (or 331-334). Pesqueras, Mexico, D.F., 9(87): 39 pages. Brocchi, P. Claus, C. 1875. Recherches sur les organes genitaux males des 1885. Neue Beitrage zur Morphologie der Crustaceen. Crustaces Decapodes. Annales des Sciences Naturelles, Arbeiten aus dem Zoologischen Institute der Universita't Zoologie et Pale'ontologie, series 6, 2(3-6): 131 pages, Wien und der Zoologischen Station in Trieste, 6(1): 108 202 figures. pages, 7 plates. Broderip, WJ. Coelho, Petronio Alves, and Marilena de Araiijo Ramos 1835. Atya. In The Penny Cyclopaedia, 3:71, unnumbered 1968. Contribution al conocimiento de los camarones figure. London. comerciales en el norte y nordeste del Brasil. Burgers, A.C.J. Documentos Te'cnicos CARPAS (Comisio'n Asesora Re- 1958. The Reactions of the Melanophores of the Fiddler gional de Pesca para el Atldntico Sudoccidental), 4(10): Crab, Uca rapax, to Crustacean Eyestalk Extracts 4 pages. and to Vertebrate Melanophorotropic Hormones. Costa, Achille Proceedings, Koninklijke Nederlandse Akademia van We- 1864. Acquisti fatti durante 1'anno 1862. Annuario del tenschappen, series C, 61:579-589, figures 1-5. Museo zoologico dell'Universita di Napoli, 2:8-94. y Burkenroad, Martin D. Coutiere, Henri 1981. The Higher Taxonomy and Evolution of Deca- 1899. Les "Alpheidae," morphologie externe et interne, poda (Crustacea). Transactions of the San Diego So- formes larvaires, bionomie. Annales des Sciences Na- ciety of Natural History, 19(17):251-268, 1 figure. turelles (Paris), Zoologie, 8(9): 559 pages, 410 figures Burukovsky, R.N. [ 1 unnumbered], 6 plates. 1974. Keys to the Shrimps, Spiny Lobsters and Homarids. 126 1906. Sur une nouvelle espece d'Alpheopsis, A. haugi prov- pages, 189 figures. Moscow: Pishchevaia promysh- enant d'un lac d'eau douce du bassin de I'Ogoue lennost [Food Industry]. [In Russian.] (Voyage de M. Haug, 1906). Bulletin du Museum Capello, F. de Brito d'Histoire Naturelle (Paris), 12:376-380, figures 1, 2. 1867. Descripc.ao de algumas especies novas ou pouco Coventry, G.A. conhecidas de Crustaceos e Arachinidos de Por- 1944. The Crustacea. In Results of the Fifth George tugal e possessoes Portuguezas do Ultramar. Me- Vanderbilt Expedition (1941) (Bahamas, Carib- morias da Academia Real das Sciencias de Lisboa, Pri- bean Sea, Panama, Galapagos Archipelago and NUMBER 364 135

Mexican Pacific Islands). Academy of Natural Sciences description des especes de ces animaux, qui vivent dans la of Philadelphia Monographs, 6:531-544. mer, sur les cotes, ou dans Us eaux douces de la France. Cowan, Charles F. xix + 446 pages, 5 tables, 56 plates. Paris: F. G. 1971. On Guerin's "Iconographie," Particularly the In- Levrault, Libraire. sects. Journal of the Society for the Bibliography of Disney, R.L.H. Natural History, 6( 1): 18-29. 1969. Phoretic Association between Simuliid Blackflies 1976. On the Disciples' Edition of Cuvier's "Regne An- and a Prawn. Transactions of the Royal Society of imal." Journal of the Society for the Bibliography of Tropical Medicine and Hygiene, 63:292. Natural History, 8(1):32-64. 1971. Association between Blackflies (Simuliidae) and Crosnier, Alain Prawns (Atyidae), with a Discussion of the Phor- 1971. Sur quelques Crustaces Decapodes ouest-africains etic Habit in Simuliids. Journal of Animal Ecology, nouveaux ou rarement signales. Bulletin du Museum 40(l):83-92, 2 figures. National d'Histoire Naturelle, series 3 (Zoologie), 1975. Speculations Regarding the Mode of Evolution of 9:569-595, 9 figures. Some Remarkable Associations between Dipt era Darnell, Rezneat M. (Cuterebridae, Simuliidae and Sphaeroceridae) 1956. Analysis of a Population of the Tropical Fresh- and Other . The Entomologist's Monthly water Shrimp Atya scabra (Leach). American Midland Magazine, 100(1319-1321):67-74. Naturalist, 55(1): 131-138, 3 figures. Doflein, F. 1962. Sex Ratios: Aquatic Animals. In Philip L. Altman 1900. Weitere Mitteilungen iiber dekapode Crustaceen and Dorothy S. Dittmer, editors, Growth Including der k. bayerischen Staatssammlungen. Sitzungsber- Reproduction and Morphological Development, pages ichte der Mathematisch-Physikalischen Classe der K6- 439-442, table 121. Washington, DC: Federation niglichen Bayerischen Akademie der Wissenschaflen zu of American Societies for Experimental Biology. Munchen, 30:125-143, 3 figures. | Biological handbooks.] Farris, James S. Da van t, Pierre 1970. Methods for Computing Wagner Trees. Systematic 1963. Clave para la identification de los camarones marinosy de Zoology, 19:83-92, 4 figures. trio con importancia economica en el oriente de Venezuela. Farris, James S., A.G. Kluge, and M.J. Eckardt 113 pages, 35 figures. Cumana, Venezuela: Insti- 1970. A Numerical Approach to Phylogenetic Sys- tuto Oceanografico, Universidad de Oriente tematics. Systematic Zoology, 19:172-189, 6 figures. (Cuadernos Oceanograficos). [Spanish text: pages 1-57, figures 1-35; English text: pages 59-113, Fausto Filho, Jose figures 1-35.] 1968. Crustaceoa decapodos de valor comercial ou util- De Haan, W. izados como alimento no nordeste Brasileiro. Bol- 1833-1850. Crustacea. In De Siebold, editor, Fauna Ja- etim Sociedade Cearense de Agronomia, Fortaleza, Ceara, ponica sive descriptio animalium, quae in itinere per 9:27, 28. Japoniam, jussu et auspiciis superiorum, qui summum in Felgenhauer, Bruce E., and Lawrence G. Abele India Batava imperium tenenl, suscepto, annis 1823- In press. Aspects of Mating Behavior in the Tropical 1830 collegit, notis observationibus et adumbrationibus Freshwater Shrimp Atya innocous (Herbst). Biotro- illustravit. xvi 4- xxxi + vii-xvii + 243 pages, plates pica. A-Q, 1-55, circular 2. Leiden: A. Arnz. Fryer, G. De Man, J.G. 1977. Studies on the Functional Morphology and Ecol- 1925. Contribution a la connaissance des Decapodes ogy of the Atyid Prawns of Dominica. Philosophical Macroures marins et fluviatiles du Congo Beige. Transactions of The Royal Society of London, Biological Annales du Muse'e du Congo Beige, series 3, section 3, Sciences, 277(952) :57-128, 121 figures. l(l):5-54, figures 1-I3d. Gerstaecker, A., and A.E. Ortmann De Montfort, Denys 1881-1901. Crustacea (ZweiteHalft: ): Die 1810. Conchyliologie systematique, et classification me'thodique Klassen und Ordnungen der Arthropoden Wissen- des coquilles. ... 676 pages. Paris: F. Schoell. schaftlich dargestellt in Wort und Bild. In H.G. De Saussure, H. Bronn, editor, Klassen und Ordnungen des Tier-Reichs, 1857. Diagnoses de quelques Crustaces nouveaux de Wissenschafilich Dargestellt in Wort und Bild, 5(2): i- 1'Amerique tropical. Revue et Magasin de Zoologie viii + 1319 pages, 128 plates. Dates of publication pure et appliquee, (2)9:501-505. are as follows; parts 1-46 by Gerstaecker, parts Desmarest, Anselme-Gaetan 47-62 by Ortmann. 1823. Malacostraces. In Dictionaire des Sciences Naturelles, Part Pages Plates Year 28:138-425; 29: 58 plates. Paris: Le Normant. 1-3 1-% 1-8 1881 1825. Considerations generates sur la classe des Crustaces, el 4-8 97-240 9-22 1882 136 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

9, 10 241-304 23-28 1883 enumerationem Insectorum, quae in museo suo adservat, 11-15 305-416 29-48 1884 examini subjecit, systematice disposuit alque descripsit. 16, 17 417-512 1886 Part 2, Pages 141 236, 17 plates. Leiden: Sump- 18, 19 513-592 49,50 1888 tibus Auctoris. 20-27 593-752 51-68 1889 Guerin-Meneville, F.E. 28 753-800 1890 1825. Le genre Atye, Atya Leach (Latr.). In M. Latreille, 29-31 801-816 69-76 1891 Le Peletier de Saint-Fargeau, Serville, and Guerin, 32-34 817-896 77-82 1892 editors, Encyclopedic Me'thodique Histoire Naturelle, 35-37 897-960 83-88 1893 10:329. Paris: M1"*veuve Agasse. 38-40 %1-976 89-% 1894 1829-1844. Iconographie du regne animal de G. Cuvier, ou 41-46 977-1056 97-108 1895 representation d'apres nature de I'une des especes Us plus 47-49 1057-1120 109-111 1898 remarquables, et souvent non figure'es, de chaque genre 50-52 1121-1168 112-116 1898 d'animaux, avec un text descriptif mis au courant de la 53-56 1169-1232 117-122 1899 science: Ouvrage pouvant servir d'atlas a tous traite's de 57-59 1233-1280 123-125 1900 zoologie. 450 plates in 45 installments (Crustacea: 61,62 1281-1319 126-128 1901 48 pages, 36 plates.]. Paris. |See Cowan, 1971, for Giebel, C.G. dates of publication.] 1875. Atya gabomnsis, neuer Krebs aus Gabon. Zeitschrift 1856 1857a. Crustaces. In Ramon de la Sagra, editor, JUT die gesammten Naturwissenschaften, 45:52-55. Histoire physique, politique et naturelle de Vile de Cuba, GifTord, C.A., and TJ. Cole 7:vii Ixviii; 8|atlas|: 3 plates. Paris: Arthus Ber- * 1970. Decapod Crustaceans of the Luquillo Experimen- trand. tal Forest. [Unpublished manuscript.] 1856- 1857b. Crustaceos. In Ramon de la Sagra, editor, Gordon, I. Historm fisica politico y natural de la isla de Cuba, 1967. Crustacea: General Considerations. In Reunion de 7:vii xxiii; 8[atlas]: 3 plates. Paris: Arthus Ber- Specialistes C.S.A. sur les Crustaces, Zanzibar trand. 1964. Memoires de I'Institut Fondamental d'Afrique Guilding, Lansdown Noire, 77:27-86, figures 1-39. 1825. An Account of Some Rare West Indian Crustacea. Gosse, Philip Henry Transactions of the Linnean Society of London, 14:334- 1851. A Naturalist's Sojourn in Jamaica, xxiv 4- 508 pages. 338. London: Longman, Brown, Green, and Long- Gundlach, Juan mans. 1887. Apuntes para la fauna Puerto-Riquena, VI: Crus- Greeff, Richard taceos. Anales de la Sociedad Espanola de Historta 1882. Die Land- und Siisswasser-Krebse der Inseln S. Natural, 16(1): 115-199. Thome und Rolas. In Eine im Winter von 1879 Haan, W. de. See De Haan auf 1880 ausgefiihrte zoologische Reise nach den Hansen, H.J. Guinea-Inseln Principe und S. Thome. Sitzungsber- 1925. Studies on Arthropoda, II: On the Comparative Morphol- ichte der Gesellschaft zur Befbrderung der gesammten ogy of the Appendages in the Arthropoda, A: Crustacea. Naturwissenschaften zu Marburg, 1882:25-37. Volume 2, 176 pages, 8 plates. Copenhagen: Gyld- 1884. Ueber die Fauna der Guinea-Inseln S. Thome endalske Boghandel. und Rolas. Sitzungsberichte der Gesellschaft zur Hart, C.W., Jr. Beforderung der gesammten Naturwissenschaften zu Mar- 1961a. Jonga, a New Genus of Freshwater Atyid Shrimps burg, 1884:41-79, 8 figures. (Decapoda, Atyidae). Notulae Naturae, Academy of Griffith, Edward Natural Sciences of Philadelphia, 342:1-3, 2 figures. 1835. A Classified Index and Synopsis of the Animal Kingdom 1961b. The Freshwater Shrimps (Atyidae and Palaemon- Arranged in Conformity with Its Organization, by the idae) of Jamaica, W.I. Proceedings of the Academy of Baron Cuvier.... cxix + 328 pages. London: Whit- Natural Sciences of Philadelphia, 113(4) :61-80, 18 taker and Co. figures. Griffith, Edward, and Edward Pidgeon 1964. A Contribution to the Limnology of Jamaica and 1833. The Classes Annelida, Crustacea, and Arachnida. Puerto Rico. Caribbean Journal of Science, 4(2,3): 331- In The Animal Kingdom Arranged in Conformity with 334, 1 figure. Its Organization by the Baron Cuvier, xin: viii + 540 1980. New Freshwater Shrimp Records for Tobago, pages, 60 plates. London: Whittaker, Treacher, West Indies, with a Summary of Records for the and Co. Lesser Antilles (Atyidae and Palaemonidae). Pro- Gronovius, Theod. ceedings of the Biological Society of Washington, 1764. Zoophylacii Gronoviani, fasciculus secundus, exhibens 93(3):845-848, 1 figure. NUMBER 364 137

Hart, C.W., Jr., and Dabney G. Hart 1957. Proposed Addition to the "Official List of Generic 1969. A Contribution to the Limnology of Dominica, Names in Zoology" of the Names of One Hundred West Indies. Proceedings of the Academy of Natural and Two Genera of (Class Crustacea, Sciences of Philadelphia, 121(4): 109-126, 5 figures. Order Decapoda), Including Proposals for the Use Hart, Dabney G., and C.W. Hart, Jr. of the Plenary Powers (a) to Validate the Emen- 1974. The Ostracod Family Entocytheridae. Academy of dation to "Gnathophyllum" of the Generic Name Natural Sciences of Philadelphia Monograph, 18: ix + "Gnatophyllum" Latreille, 1814, and (b) to Vali- 239 pages, 49 figures, 52 plates. date the Family-Group Names "Hippolytidae" Hemming, A.F. Bate, 1888, and "Eugonatonotidae" Chace, 1937. 1957. Opinion 470: Addition to the "Official List of In A.F. Hemming, Opinion 470..., pages 162-188. Generic Names in Zoology" of the Names of One [See Hemming, 1957.] Hundred and Two Genera of Caridea (Class Crus- 1959. The Crustacea Decapoda of Suriname (Dutch tacea, Order Decapoda) and Use of the Plenary Guiana). Zoologische Verhandelingen Uitgegeven door Powers for Various Purposes in Connection There- het Rijksmuseum van Natuurlijke Historie te Leiden, 44: with. In Opinions and Declarations Rendered by the 2% pages, 68 figures, 16 plates. International Commission on Zoological Nomenclature, 1963. Two New Species of Fresh-Water Shrimp (Crus- 16:129-202. London. tacea Decapoda) from the West Indies. Proceedings, Herbst, Johann F.W. Koninklijke Nederlandse Akademie van Wetenschappen, 1791-1796. Versuch einer Naturgeschichte der Krabben und section C, 66(l):61-69, 3 figures. Krebse nebst einer systematischen Beschreibung ihrer ver- 1966. The R/V Pillsbury Deep-Sea Biological Expedition schiedenen Arten, 2: viii + 266 pages, plates 22-46. to the Gulf of Guinea, 1964-65, 11: The Fresh- Berlin: Gottlieb August Lange. water Shrimps of the Island of Annobon, West Hickson, Sydney J. Africa. Studies in Tropical Oceanography, 4(1):224- 1889. A Naturalist in North Celebes, xv + 392 pages. Lon- 239, 5 figures. don: John Murray. 1969. Etudes hydrobiologiques en Nouvelle-Caledonie Hobbs, Horton H., Jr. (Mission 1965 du Premier Institut de Zoologie de 1971. The Entocytherid Ostracods of Mexico and Cuba. l'Universite de Vienne (suite)), DC: The Freshwa- Smithsonian Contributions to Zoology, 81: 55 pages, 31 ter Shrimps (Crustacea, Decapoda, Natantia) of figures. New Caledonia. Cahiers O.R.S.T.O.M., Serie Hydro- 1980. Atya gabonensis (Decapoda, Atyidae) in the Western biologie, 3(2):87-108, 4 figures. Hemisphere. Crustaceana, 38(1):111. 1974. Subterranean Crustacea Decapoda Macrura Hobbs, Horton H., Jr., H.H. Hobbs III, and Margaret A. Collected by Mr. L. Botosaneanu during the Daniel 1973 Cuban-Roumanian Biospeological Expedi- 1977. A Review of the Troglobitic Decapod Crustaceans tion to Cuba. International Journal of Speleology, of the Americas. Smithsonian Contributions to Zoology, 6(1974):231-242, 2 figures. 224: v + 183 pages, 70 figures. 1977. On Some Freshwater and Terrestrial Crustacea De- Holthuis, L.B. capoda from Cuba. In Re'sultats des expeditions bio- 1951. The Caridean Crustacea of Tropical West Africa. speologiques Cubano-Roumaines a Cuba, pages 271— Atlantide Report, 2:7-187, figures 1-34. 275, 1 figure. Bucharest: Academiei Republicii 1955a. Proposed Addition to the "Official List of Generic Socialiste Romania. Names in Zoology" of the Names of One Hundred 1980. Shrimps and Prawns of the World: An Annotated and Two Genera of Caridea (Class Crustacea, Catalogue of Species of Interest to Fisheries. FAO Order Decapoda), Including Proposals for the Use Fisheries Synopsis, 125: xvii + 271 pages. [Volume of the Plenary Powers (a) to Validate the Emen- 1 of FAO Species Catalogue.] dation to "Gnathophyllum" of the Generic Name Holthuis, L.B., and H. Rosa, Jr. "Gnatophyllum" Latreille, 1814, and (b) to Vali- 1965. List of Species of Shrimps and Prawns of Eco- date the Family-Group Names "Hippolytidae" nomic Value. FAO Fisheries Technical Paper, 52: iii Bate, 1888 and "Eugonatonotidae" Chace, 1937. + 21 pages. Bulletin of Zoological Nomenclature, 11:204-228. Hunte, Wayne 1955b. The Recent Genera of the Caridean and Steno- 1975. Atya lanipes Holthuis, 1963, in Jamaica, Including podidean Shrimps (Class Crustacea, Order Deca- Taxonomic Notes and Description of the First poda, Supersection Natantia) with Keys for Their Larval State (Decapoda, Atyidae). Crustaceana, Determination. Zoologische Verhandelingen Uitgegeven 28(l):66-72, 3 figures. door het Rijksmuseum van Natuurlijke Historie te Leiden, 1977. Laboratory Rearing of the Atyid Shrimps Atya 26:157 pages, 105 figures. innocous (Herbst) and Micratya poeyi (Guerin-Mene- 138 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

ville) (Decapoda, Atyidae). Aquaculturt, 11(4):373- W. and Lorna R. Levi, volume 3, xi + 523 pages, 378. 18 figures. New York: Interscience Publishers 1978. The Distribution of Freshwater Shrimps (Atyidae (John Wiley & Sons). and Palaemonidae) in Jamaica. Zoological Journal Kingsley, J.S. of the Linnean Society of London, 64(2): 135-150, 7 1878a. Notes on the North American Caridea in the figures. Museum of the Peabody Academy of Science at 1979a. The Complete Larval Development of the Fresh- Salem, Mass. Proceedings of the Academy of Natural water Shrimp Micratya poeyi (Guerin-Meneville) Sciences of Philadelphia, 1878:89-98. Reared in the Laboratory (Decapoda, Atyidae). 1878b. List of the North American Crustacea Belonging Crustaceana, supplement, 5:151-166, 8 figures. to the Sub-order Caridea. Bulletin of the Essex 1979b. The Complete Larval Development of the Fresh- Institute, 10(4-6): 53-71. water Shrimp Atya innocous (Herbst) Reared in the Koelbel, Carl Laboratory (Decapoda, Atyidae). Crustaceana, sup- 1884. Carcinologisches. Silzungsberichte der Kaiser lichen plement, 5:231-242, 5 figures. Akademieder Wissenschaflen Wien, 90(1):312-323, 3 1979c. The Rediscovery of the Freshwater Shrimp Macro- plates. brachium crenulatum in Jamaica. In P. Wagenaar Latreille, M. Hummelinck and L.J. Van der Steen, editors, 1817. Les Crustaces, Les Arachnides et Les Insectes. In Studies on the Fauna of Curacao and Other Caribbean G. Cuvier, editor, Le Re'gne Animal, distribue d'apres Islands, 58:69-73, figure 58. Utrecht: Foundation son organisation, pour servir de base a I'histoire naturelle for Scientific Research in Surinam and the Neth- des animaux et d'introduction a I'anatomie compare'e, 3: erlands Antilles. xxix + 653 pages. Paris. International Commission on Zoological Nomenclature 1837. Mollusca-Annelides-Crustacea-and Arachnides. In (ICZN) G. Cuvier, editor, The Animal Kingdom, Arranged 1954. Opinion 260. Rejection for Nomenclatorial Pur- According to Its Organization, Serving As a Foundation poses of the Work by Meuschen (F.C.) Issued in for the Natural History of Animals, and an Introduction 1778 under the Title Museum Gronovianum. In to Comparative Anatomy, 3: xxiv + 199 pages. Lon- Opinions and Declarations Rendered by the International don: G. Henderson. Commission on Zoological Nomenclature, 5(21): Leach, William Elford 265-280, 2 plates. 1815. A Tabular View of the External Characters of Irvine, F.R. Four Classes of Animals, Which Linne Arranged 1947. Crustaceans, Turtles, Cetaceans, etc. In F.R. Ir- under Insecta, with the Distribution of the Genera vine, The Fishes and Fisheries of the Gold Coast, pages Composing Three of These Classes into Orders, 283-320, figures 191-217. London: Crown Agents &c. and Descriptions of Several New Genera and for the Colonies. [Printed in Great Britain at the Species. Transactions of the Linnean Society of London, University Press, Cambridge (Brooke Crutchley, 11:306-400. University Printer).] 1816 [1824]. Atya. In Encyclopaedia Britannica, supplement Johnston, Harry to the fourth, fifth, and sixth editions, 1(1):421, 1906. Liberia. Volume 2, xvi + 521-1183 pages, 426 plate 21. Edinburgh: Archibald Constable and figures. London: Hutchinson & Co. Company. [Reprinted 1824.] Jonston, J. 1817. The Zoological Miscellany. Volume III, v + 151 * 1660. Beschryvingh van dt Natuur der Bloedlose Waterdieren in pages, 149 plates. London: R.P. Nodder. vier Boeken. 56 pages, 20 plates. Amsterdam: I.I. Lebour, Marie V. Schipper. 1959. The Larval Decapod Crustacea of Tropical West Jonstonus, Joannes Africa. Atlantide Report, 5:119-143, 21 figures. * 1650. Historiae naturalis de exanguibus aquaticis. Book 4, 78 Lemasson, J. pages, 20 plates. Frankfurt: M. Merian. 1973. Chronique Pisicole. Bois et Forets des Tropiques, * 1657. Historiae naturalis de exanguibus aquaticis. Book 4, 58 152:61-70. pages, 20 plates. Amsterdam: JJ. Schipper. [Reis- Lemos de Castro, Alceu sued in 1660 with new title page.] 1962. Sobre os crustaceos referidos por Marcgrave em 1665. Historiae naturalis de exanguibus aquaticis. Book 4, 60 sua "Historia Naturalis Brasiliae" (1648). Arquivos pages, 20 plates. Amsterdam: JJ. Schipper. do Museu Nacional (Brazil), 52:37-51, 29 figures. * 1767. Historiae naturalis de exanguibus aquaticis. Book 4, 78 Leveque, Christian pages, 20 plates. Heilbrunn: F.J. Eckebrecht. 1974. Crevettes d'eau douce de la Guadeloupe (Atyidae Kaestner, Alfred et Palaemonidae). Cahiers O.R.S.T.O.M. (I'Officede 1970. Invertebrate Zoology, Crustacea. Translated by Herbert la Recherche Scientifique et Technique Outre-Mer), Serie NUMBER 364 139

Hydrobiologie, 8(l):41-49, 1 plate, 5 figures. editor, Zoophylacium Gronovianum, exhibens Animalia Lewis, D.J., R.H.L. Disney, and R.W. Crosskey Quadrupeda, Amphibia, Pisces, Insecta, Vermes, Mol- 1969 [ 1968]. A New Phoretic Species of Simultum (Dipt., lusca, Testacea, et Zoophyta, quae in museo suo adservavit, Simuliidae) from West Cameroon, with Taxo- examini subjecit, systematice disposuit atque descripsit, 19 nomic Notes on Allied Forms. Bulletin of Entomolog- unnumbered pages. Leiden: T. Haak & Co., S. ical Research, 59(2).229-239, 20 figures. and J. Luchtmans. Lucas, Pierre Hippolyte Milne-Edwards, Alphonse 1833-1834. Atye, Atya. In Dictionnaire pittoresque d'histoire 1864. Revision des Crustaces Macroures de la famille naturelle et des phenomenes de la nature, 1:333. Paris: des Atyoidees. Annales de la Socie'te Entomologique de Au Bureau de Souscription. France, series 4, 4:145-152, plate 3. 1840 [1842, 1851]. Histoire Naturelle des Crustaces, des Ar- Milne Edwards, Henri achnides et des Myriapodes. 601 pages, 46 plates. 1816. Les Crustaces. In Georges Cuvier, editor, Le Regne Paris: P. Dumenil. [Reprinted in 1842 and 1851.] animal distribue d'apres son organisation, pour servir de Man, J.G., de. See De Man base a I'histoire naturelle des animaux, et d'introduction a Manning, Raymond B., and L.B. Holthuis I'anatomie comparee, vm: 278 pages, plate 51. Paris: 1981. West African Brachyuran (Crustacea: De- Fortin, Masson et C, Libraires. capoda). Smithsonian Contributions to Zoology, 306: xii 1836-1844. Les Crustaces. In Georges Cuvier, editor, Le + 379 pages, 87 figures. regne animal distribue'd'apres son organisation, pour servir Marcgrave, G. de base a I'histoire naturelle des animaux, et d'introduction 1648. Historiae rerum naturalium Brasiliae, libri octo: a I'anatomie comparee, 17: 278 pages; 18, 80 plates. Quorum tres priores agunt de plantis, quartus de Paris: Fortin, Masson et C16, Libraires. [For a piscibus, quintus de avibus, sextus de quadrupe- collation of publication dates, see Cowan, 1976:60, dibus, et serpentibus, septimus de insectis, octavus or Manning and Holthuis, 1981:369.] de ipsa regione, et illius incolis, cum appendice de 1837. Histoire naturelle des Crustaces, comprenant I'anatomie, la Tapuyis, et Chilensibus. In G. Piso and G. Marc- physiologie et la classification de ces animaux. Volume graf de Liebstad, editors, Historia naturalis Brasiliae, 2, 532 pages; atlas, 32 pages, plates 1-14, 14bis, auspicio et beneficio illustriss, I: Mauritii Com. Nassau 15-25, 25bis, 26-42. Paris: Libraire encyclo- illius provincial et maris summi praefecti adornata, in qua pedique de Roret. non tantum plantae et animalia, sed et indigenarum morbi, 1838. [Contribution.] In J.B.P.A. de Lamarck, editor, ingenia et mores describuntur et iconibus supra quingentas Histoire naturelle des animaux sans vertebres, second illustrantur, part 2, 293 pages, illustrated. edition, 2: 699 pages. Paris: J.B. Bailliere. 1942. Historia Natural do Brasil. Translated by Jose Pro- 1839. [Contribution.] In J.B.P.A. de Lamarck, editor, copio de Magalhaes, iv + 297 + civ pages, illus- Histoire naturelle des animaux sans vertebres, third edi- trated. Sao Paulo: Imprensa Oficial do Estado. tion, 2: 436 pages. Brussels: Meline, Cans et Com- Martens, Eduard von. See Von Martens pagnie. Mendez G., Matilde Mocquard, F. 1981. Claves de identificacion y distribucion de Ios lan- 1883. Recherches anatomiques sur 1'estomac des Crus- gostinos y camarones (Crustacea: Decapoda) del taces podophthalmaires. Annales des Sciences Natu- mar y rios de la costa del Peru. Instituto del Mar del relles (Paris), Zoologie, series 6, 16(1): 311 pages, 11 Peru, Boletin, 5: 170 pages, 57 plates. plates. Meuschen, F.C. Monod, Th. 1778. Mvsevm Gronovianvm, sive index rervm natvralivm tarn 1928. L'induslrie des peches au Cameroun: Mission Monod Mammalivm, Amphibiorvm, Piscivm, Insectorvm, Con- (1925-1926) Cameroun. Volume 1,504 + 5 unnum- chyliorvm, Zoophytorvm, Plantarvm, et Mineralivm bered pages, % figures, 25 plates. Paris: Societe exqvisitissimorvm qvam arte factarvm nonnvllarvm, inter d'editions Geographiques, Maritimes et Colonaise. qvae emmet herbarivs siccvs plantarvm a Tovmefortio 1933. Sur quelques Crustaces de l'Afrique Occidentale Claitonio Linnaeo aliisqve botanicis collectarvm, qvae om- (Liste des Decapodes mauritaniens et des Xan- nia mvlta cvra et magnis svmtibvs sibi comparavit vir thides ouest-africains). Bulletin du Comite d'Etudes amplissimvs & celeberrimvs Lavr. Theod. Gronovivs Historiques et Scientifiques de l'Afrique Occidentale Fran- J. V.D. Civitatis Lvgdvno Batavae Senator et Scabinvs qaise, 15(2-3):456-548, 26 figures. Societatis Regiae Londinensis Brasilaeensis et Hollandiae 1967. Crevettes et crabes de la cote occidentale qvae Harlemi est aliorvmque Socivs &c. &c, i-vi + 231 d'Afrique. In Reunion de specialistes C.S.A. sur pages, 1 plate. Rotterdam. les Crustaces. Memoires de I'lnstitut Fondamental * 1781. Index, continens nomina generica specierum pro- d'Afrique Noire, 77:103-234, 26 plates. pria, trivialia ut et synonyma. In L.T. Gronovius, 1977. Sur quelques Crustaces Decapodes africains 140 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

(Sahel, Soudan). Bulletin du Museum National 1897. Os Camaroes da agua doce da America do Sul. d'Histoire Naturelle, 3rd scries, 500:1201-1236, 131 Revista do Museu Paulista, 2:173-216, plate 1. figures. Osorio, Balthazar 1980. Decapodes. In J.R. Durand and C. Leveque, edi- 1887. Liste des Crustaces des possessions Portugaises tors, Flore et fount aquatiques dt I'Afrique Sahe'lo-Sou- D'Afrique occidentale dans les collections du Mu- danienne, 1:369-389, figures 1-35. Paris: OR- seum D'Histoire Naturelle de Lisbonne. Jornal de STOM. Sciencias Mathemalicas, Physicas, e Naturaes (Lisboa), Montfort, Denys de. See De Montfort 11:220-231. Motwani, M.P., and Y. Kanwai 1888. Liste des Crustaces des possessions Portugaises 1970. Fish and Fisheries of the Coffer-dammed Right D'Afrique occidentale dans les collections du Mu- Channel of the River Niger at Kainji. In S.A. seum D'Histoire Naturelle de Lisbonne. Jornal de Visser, editor, Kainji, a Nigerian Man-made Lake. Sciencias Mathematicas, Physicas, e Naturaes (Lisboa), In Kainji Lake Studies, Volume I: Ecology, pages 12:186-191. 27-48, 3 figures, 2 plates. Nigerian Institute of 1889. Nouvelle contribution pour la connaissance de la Social and Economic Research. faune carcinologique des lies Saint Thome et du Nagamine, Claude M., and Allen W. Knight Prince. Jornal de Sciencias Mathematicas, Physicas, e 1980. Development, Maturation, and Function of some Naturaes (Lisboa), series 2, 1:129-139. Sexually Dimorphic Structures of the Malaysian 1891a. Note sur quelques especes de Crustaces des lies S. Prawn, Macrobrachium rosenbergii (De Man) (Deca- Thome, du Prince et Ilheo das Rolas. Journal de poda, Palaemonidae). Crustaceana, 39(2): 141-152, Sciencias Mathematicas, Physicas, e Naturaes (Lisboa), 4 figures. series 2, 2(5):45-49. Newport, G. 1891b. Note sur quelques especes de Crustaces de ltle 1847. Note on the Genus Atya of Leach, with Descrip- Saint Thome, Hot das Rolas et Angola. Jornal de tions of Four Apparently New Species in the Sciencias Mathematicas, Physicas, e Naturaes (Lisboa), Cabinets of the British Museum. Annals and Mag- series 2, 2(6): 140, 141. azine of Natural History, 19:158-160, plate 8: figure 1892. Nova contribuicao para a fauna carcinologica da 1. ilha de S. Thome. Jornal de Sciencias Mathematicas, Nobili, Giuseppe Physicas, e Naturaes (Lisboa), series 2, 2:199-204. 1897. Decapodi e Stomatopodi raccolti dal Dr. Enrico 1895a. Crustaceos da Ilha d'Anno Bom. Journal de Sciencias Festa nel Darien, a Curacao, La Guayra, Porto Mathematicas, Physicas, e Naturaes (Lisboa), series 2, Cabello, Colon, Panama, ecc. Bollettino dei Musei di 3(12): 248-250. Zoologia ed Anatomia comparata delta R. Universita di 1895b. Crustaceos da Ilha Principe. Jornal de Sciencias Torino, 12(280): 1-8. Mathematicas, Physicas, e Naturaes (Lisboa), series 2, Odum, Howard T. 3(12):251. 1970. Rain Forest Structure and Mineral-Cycling Ho- 1898. Da distribuicao geographica dos Peixes e Crusta- meostasis. In Howard T. Odum and Robert F. ceos colhidos nas possessoes Portuguezas d'Africa Pigeon, editors, A Tropical Rain Forest, a Study of occidental e existentes no Museu Nacional de Irradiation and Ecology at El Verde, Puerto Rico, pages Lisboa. Jornal de Sciencias Mathematicas, Physicas, e H-3-H-52,47 figures. Washington, D. C: Division Naturaes (Lisboa), series 2, 5(19): 185-202. of Technical Information, U.S. Atomic Energy 1905. Breve noticia acerca de alguns Peixes e Curstaceos Commission. colhidos nas possessoes Portuguezas da Africa oc- Oliveira, Lejeune P.H. de cidental. Jornal de Sciencias Mathematicas, Physicas, e 1945. Verificacao da existencia de Atya scabra Leach, Naturaes (Lisboa), series 2, 7(26):97-102. camarao d'agua doce da fami'lia Atyidae, Crustacea, 1906. Uma nova lista de Crustaceos africanos. Jornal de no nordeste do Brasil. Memo'rias do Instiluto Oswaldo Sciencias Mathematicas, Physicas, e Naturaes (Lisboa), Cruz, 43(2): 177-190, plates 1, 2. series 2, 7(27): 149-150. Ortmann, A.E. Parodiz, Juan Jose 1890. Die Decapoden-Krebse des Strassburger Mu- 1960. Atyidae neotropicales (Crustacea-Natantia-Cari- seums, I. Theil. Zoologische Jahrbiicher Abtheilung fur dea) en la coleccion del Museo Carnegie. Neotro- Systematik, Geographie und Biologie der Thiere, pica, 6(20): 38-40, figures 1-7. 5:437-542, plates 36, 37. Pearse, A.S. 1895. A Study of the Systematic and Geographical Dis- 1915. An Account of the Crustacea Collected by the tribution of the Decapod Family Atyidae King- Walker Expedition to Santa Marta, Colombia. sley. Proceedings of the Academy of Natural Sciences of Proceedings of the United States National Museum, Philadelphia, 46 (1894):397-416. 49(2123) :531-556, plates 70-73. NUMBER 364 141

Peck, Stewart B. bered pages, 205 figures. Zaria, Northern Nigeria: 1975. The Invertebrate Fauna of Tropical American Ministry of Agriculture Northern Nigeria. Caves, Part III: Jamaica, an Introduction. Inter- Rioja Lo Bianco, Enrique, Manuel Ruiz Oronoz, and Ig- national Journal of Speleology, 7:303-326, 8 figures. nacio Larios Rodriguez Pericchi Lopez, Juan J. 1955. Tratado elemental de zoologia. Third edition, xv + 737 1965. La industna del camaron en Venezuela: Explotacion y pages, 714 figures. Mexico City: Editorial Porrua, pTocesamiento del camaron. 4 unnumbered pages, SA. pages 1-69, 7 unnumbered pages, maps 1-4, Rodriguez, Gilberto graphs l-7a. Caracas: Corporacion Venezolana 1981. Decapoda. In S.H. Hurlbert, G. Rodriguez, and de Fomento. N.D. Santos, editors, Aquatic Biota of Tropical South Pesta, Otto America, Part I: Arthropods. San Diego, Calfifornia: 1931. Crustacea Decapoda aus Costa Rica. In Ergebnisse San Diego Sate University. der Osterreichischen Biologischen Costa-Rica-Ex- Roux, Jean pedition 1930, I. Teil. Annalen des Naturhistorischen 1926a. An Account of Australian Atyidae. Records of the Museums in Wien, 45:173-181, plates 5, 6. Australian Museum, 15(3):237-254. Pinchon, R. 1926b. Crust aces decapodes d'eau douce de la Nouvelle- 1967. Quelques aspects de la nature aux Antilles. 254 pages, Caledonie. In Fritz Sarasin and J. Roux, editors, 46 plates. Fort-de-France, Martinique: R. Pin- Nova Caledonia, Zoologie, 4(2): 181-240, 56 figures. chon. Munich: C.W. Kreidel's Verlag. Piso, G. 1932. Siisswassermacruren der Deutschen Limnolo- 1658. Historiae naturalis & medicae Indiae occidentalis gischen Sunda-Expedition. Archivfur Hydrobiologie, libri quinque. In De Indiae utriusque re naturali et supplement, 11:563-574, 1 figure. medica libri quatuordecim, quorum contenta pagina sequens Roux, Polydore exhibet, pages 1-327 + 5 unnumbered pages and 1831. Memoire sur la classification des Crustaces de la tribu des figures. Amsterdam: Elsevier. Salicoques. Pages 1-39 + 3 unnumbered pages and 1957. Historia natural e medica da India ocidental, em table. Marseille: Imprimerie Militaire de Dufort cinco livros. In Coleqao de Obras Raras, 5: xix + 685 Cadet. pages, figures. Rio de Janeiro. Rutherford, T.C. Pocock, R.I. 1971. Fresh-Water Shrimps in the Area of Cape Coast, 1889. Contributions to Our Knowledge of the Crustacea Ghana. Ghana Journal of Science, 11(2):87-91, 9 of Dominica. Annals and Magazine of Natural History, figures. series 6, 3:6-22, figures 1-8. Ruysch, H. 1894. Contributions to Our Knowledge of the Arthropod * 1718. Theatri universalis animalium pars quarta sive historiae Fauna of the West Indies, Part I: Scorpiones and naturalis de exanguibus aquaticis libri IV: Cum enumer- Pedipalpi; with a Supplementary Note upon the atione morborum, quibus medicamina ex his animalibus Freshwater Decapoda of St. Vincent. Journal of the petuntur, ac notitia animalium, ex quibus vicissim remedia Linnean Society of London, 24:374-409, plates 29, 30. praestantissima possunt capi. 58 pages, 20 plates. Am- Powell, C.B. sterdam. 1979. Three Alpheid Shrimps of a New Genus from Saussure, H. de. See De Saussure West African Fresh and Brackish Waters: Tax- onomy and Ecological Zonation (Crustacea De- Sawaya, Paulo capoda Natantia). Revue de Zoologie Africaine, 1942. Comentarios por os Crustaceos. In Jorge Marc- 93(1): 116-150, 8 figures. grave, Historia natural do Brasil, translated by Jose Rathbun, Mary J. Procopio de Magalhaes, LXI-LXV. Sao Paulo: Im- 1897. List of the Decapod Crustacea of Jamaica. Annals prensa Oficial do Estado. of the Institute of Jamaica, 1(1): 45 pages. Scelzo, M.A. 1900. The Decapod Crustaceans of West Africa. Proceed- 1974. Tecnicas para la produccion y obtencion de larvas, ings of the United States National Museum, postlarvas, y juveniles en el cultivo de crustaceos 22(1199):271-316, 2 figures. en America Latina. In Simposio FAO/CARPASsobre 1901. The Brachyura and Macrura of Porto Rico. United acuicultura en America Latina, CARPAS/6/74/SR, States Fish Commission Bulletin for 1900 [1902], 20:1-127, 129-137 [index], 24 figures, 1 plate, 6:1-17. frontispiece. [Preprint.] Schmitt, Waldo L. Reed, William (with contributions by John Burchard, A.J. 1935. Crustacea Macrura and Anomura of Porto Rico Hopson, Jonathan Jenness, and Ibrahim Yaro). and the Virgin Islands. Scientific Survey of Porto Rico 1967. Fish and Fisheries of Northern Nigeria. 226 + unnum- and the Virgin Islands, 15(2): 125-227, figures 1-80. 142 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY

Sharp, Benj. Velez, Manuel J., Jr. 1893. Catalogue of the Crustaceans in the Museum of 1967. Checklist of the Terrestrial and Freshwater Deca- the Academy of Natural Sciences of Philadelphia. pods of Puerto Rico. Caribbean Journal of Science, Proceedings of the Academy of Natural Sciences of Phil- 7(l-2):41-44. adelphia, 1893:104-127. Vilanova y Piera, J. Sheldon, J.M. Arms * 1875. Articulados, la creacion: Historia natural escrita por una 1905. Guide to the Invertebrates of the Synoptic Collection in the sociedadde Naturalistas. Volume 6, i-xii + 656 pages, Museum of the Boston Society of Natural History, v + 748 figures, 12 plates. Barcelona: Montaner y 505 pages. Boston: Boston Society of Natural His- Simon. tory. Villalobos F., Alejandro Smith, Sidney I. 1943. Estudio morfologico de la Atya scabra (Crust. De- 1871. List of the Crustacea Collected by J.A. McNiel in cap.). 70 pages, 22 plates. Tesis, Universidad Na- Central America. Report of the Peabody Academy of cional Autonoma de Mexico, Mexico City. Science, 1869:87-98. 1956. Contribucion al conocimiento de los Atyidae de Solar C, Enrique M. del Mexico, I: Una nueva especie de Atya de la ver- 1972. Addenda al catalogo de Crustaceos del Peru. Insti- tiente del Pacifico del Estado de Michoacan. An- tute del Mar del Peru, Informe, 38:1-21. ales del Institute de Biologia, Universidad Nacionat Au- Stebbing, Thomas R.R. tonoma de Mexico, 26(2):459-475, plates 1 -6. 1893. A History of Crustacea: Recent Malacostraca. xvii + 466 1959. Contribucion al conocimiento de los Atyidae de pages, 19 plates. New York: D. Appleton and Mexico, II (Crustacea, Decapoda): Estudio de Company. algunas especies del genero Potimirim (= Orlman- Stimpson, William 1857. On the Crustacea and Echinodermata of the Pa- nia), con descripcion de una especie nueva de cific Shores of . Boston Journal of Brasil. Anales del Institute de Biologia, Universidad Nacional Autonoma de Mexico, 30(l-2):269-330, 12 Natural History Containing Papers and Communications Read to the Boston Society of Natural History, figures. 6(4):444-532, plates 18-23. Villamil, Johnny, and Richard G. Clements Straskraba, M. 1976. Some Aspects of the Ecology of the Fresh Water Shrimps 1969. Lista de los Crustaceos dulceacuicolas de Cuba y in the Upper Espiritu Santo River at El Verde, Puerto sus relaciones zoogeograficas. Academia de Ciencias Rico, vii + 62 pages, 12+1 figures. Puerto Rico de Cuba, Institute de Biologia, Serie Biolo'gica, 8:1-37. Nuclear Center, University of Puerto Rico Tesch,J.J. (United States Energy Research and Development 1914. Crustacea. In H.D. Benjamins and J.F. Snelleman, Administration, publication 206). editors, Encyclopaedie van Nederlandsch West-Indie, Vogel, H., and W. Crewe pages 246-250. 1965. Beobachtungen iiber die Lungenegel-Infektion in Thompson, D'Arcy Wentworth Kamerun (Westafrika). Zeitschnft fur Tropenmedizin 1901. A Catalogue of Crustacea and of Pycnogonida Contained und Parasitologie, 16(2): 109-125, 11 figures. in the Museum of University College, Dundee. 56 pages. Voigt, F.S. Dundee. [Printed for the Museum.] 1836. Die Anneliden, Crustaceen, Arachniden und die Torralbas, Federico ungefliigelten Insekten. In Cuvier, editor, Das 1917. Contribucion al estudio de los crustaceos de Cuba: Thierreich, geordnet nach seiner Organization: Als Grun- Notas del Dr. Juan Gundlach, fl896 compiladas dlage der Naturgeschichte der Thiere und Einleitung in y completadas por el Dr. Jose I. Torralbas, fl9O3. die vergleichende Anatomie, 4: xiv + 516 pages. Anales de la Academia de Ciencias Medicos, Fisicas y Leipzig. Naturales de la Habana, 53:543-624, 73 figures. Von Martens, Eduard Upatham, E.S., and R.F. Sturrock 1868. Ueber einige ostasiatische Siisswasserthiere. Archiv 1973. Field Investigations on the Effect of Other Aquatic fur Naturgeschichte, 34(1): 64 pages. Animals on the Infection of Biomphalaria glabrata 1872. Ueber Cubanische Crustaceen nach den Sa- by Schistosoma mansoni Miracidia. Journal of Parasi- mmlungen Dr. J. Gundlach's. Archiv fur Naturges- tology, 59(3):448-453, 2 figures. chichte, 38(1):77-147, plates 4, 5. Valdes Ragues, Pedro White, Adam 1909. Clasificacion "Gundlach" de Crustaceos cubanos. 1847. List of the Specimens of Crustacea in the Collection of the In Mis Trabajos Academicos, pages 163-187. Ha- British Museum, viii + 143 pages. London: Edward vana: Imprenta de Cuba Intelectual. Newman. NUMBER 364 143

Wiegmann, A.F. A. Fish and Fisheries of Northern Nigeria, pages 203-212. 1836. Beschreibung einiger neuen Crustaceen des Ber- Zaria, Northern Nigeria: Ministry of Agriculture, liner Museums aus Mexiko und Brasilien. Archiv Northern Nigeria. fur Naturgeschichte, 2 (1): 145-151. Young, Charles G. Yaro, Ibrahim 1900. The Stalk-eyed Crustacea of British Guiana, West Indies, 1967. Common Names of Local Fish in English, Hausa, and Bermuda, xix + 514 pages. London: John M. Nupe, Ijaw and Kanuri. In William Reed, editor, Watkins.

REQUIREMENTS FOR SMITHSONIAN SERIES PUBLICATION Manuscripts intended for series publication receive substantive review within their originating Smithsonian museums or offices and are submitted to the Smithsonian Institution Press with approval of the appropriate museum authority on Form SI-36. Requests for special treatment—use of color, foldouts, casebound covers, etc.—require, on the same form, the added approval of designated committees or museum directors. Review of manuscripts and art by the Press for requirements of series format and style, completeness and clarity of copy, and arrangement of all material, as outlined below, will govern, within the judgment of the Press, acceptance or rejection of the manuscripts and art. Copy must be typewritten, double-spaced, on one side of standard white bond paper, with iyA" margins, submitted as ribbon copy (not carbon or xerox), in loose sheets (not stapled or bound), and accompanied by original art. Minimum acceptable length is 30 pages. Front matter (preceding the text) should include: title page with only title and author and no other information, abstract page with author/title/series/etc., following the establish- ed format, table of contents with indents reflecting the heads and structure of the paper. First page of text should carry the title and author at the top of the page and an unnum- bered footnote at the bottom consisting of author's name and professional mailing address. Center heads of whatever level should be typed with initial caps of major words, with extra space above and below the head, but with no other preparation (such as all caps or underline). Run-in paragraph heads should use period/dashes or colons as necessary. Tabulations within text (lists of data, often in parallel columns) can be typed on the text page where they occur, but they should not contain rules or formal, numbered table heads. Formal tables (numbered, with table heads, boxheads, stubs, rules) should be sub- mitted as camera copy, but the author must contact the series section of the Press for edito- rial attention and preparation assistance before final typing of this matter. Taxonomic keys in natural history papers should use the alined-couplet form in the zoology and paleobiology series and the multi-level indent form in the botany series. If cross-referencing is required between key and text, do not include page references within the key, but number the keyed-out taxa with their corresponding heads in the text. Synonymy in the zoology and paleobiology series must use the short form (taxon, author, year:page), with a full reference at the end of the paper under "Literature Cited." For the botany series, the long form (taxon, author, abbreviated journal or book title, volume, page, year, with no reference in the "Literature Cited") is optional. Footnotes, when few in number, whether annotative or bibliographic, should be typed at the bottom of the text page on which the reference occurs. Extensive notes must appear at the end of the text in a notes section. If bibliographic footnotes are required, use the short form (author/brief title/page) with the full reference in the bibliography. Text-reference system (author/year/page within the text, with the full reference in a "Literature Cited" at the end of the text) must be used in place of bibliographic footnotes in all scientific series and is strongly recommended in the history and technology series: "(Jones, 1910:122)" or ".. . . Jones (1910:122)." Bibliography, depending upon use, is termed "References," "Selected References," or "Literature Cited." Spell out book, journal, and article titles, using initial caps in all major words. For capitalization of titles in foreign languages, follow the national practice of each language. Underline (for italics) book and journal titles. Use the colon-parentheses system for volume/number/page citations: "10(2):5-9." For alinement and arrangement of elements, follow the format of the series for which the manuscript is intended. Legends for illustrations must not be attached to the art nor included within the text but must be submitted at the end of the manuscript—with as many legends typed, double- spaced, to a page as convenient. Illustrations must not be included within the manuscript but must be submitted sepa- rately as original art (not copies). All illustrations (photographs, line drawings, maps, etc.) can be intermixed throughout the printed text. They should be termed Figures and should be numbered consecutively. If several "figures" are treated as components of a single larger figure, they should be designated by lowercase italic letters (underlined in copy) on the illus- tration, in the legend, and in text references: "Figure 9b.." If illustrations are intended to be printed separately on coated stock following the text, they should be termed Plates and any components should be lettered as in figures: "Plate 9.b_." Keys to any symbols within an illustration should appear on the art and not in the legend. A few points of style: (1) Do not use periods after such abbreviations as "mm, ft, yds, USNM, NNE, AM, BC." (2) Use hyphens in spelled-out fractions: "two-thirds." (3) Spell out numbers "one" through "nine" in expository text, but use numerals in all other cases if possible. (4) Use the metric system of measurement, where possible, instead of the English system. (5) Use the decimal system, where possible, in place of fractions. (6) Use day/month/year sequence for dates: "9 April 1976." (7) For months in tabular list- ings or data sections, use three-letter abbreviations with no periods: "Jan, Mar, Jun," etc. Arrange and paginate sequentially EVERY sheet of manuscript—including ALL front matter and ALL legends, etc., at the back of the text—in the following order: (1) title page, (2) abstract, (3) table of contents, (4) foreword and/or preface, (5) text, (6) appendixes, (7) notes, (8) glossary, (9) bibliography, (10) index, (11) legends. «•

,111/.,